modling_forDSM
gaohong
2019年6月30日
设置工作路径，加载包
library(rgdal)## Warning: package 'rgdal' was built under R version 3.5.3## Loading required package: sp## Warning: package 'sp' was built under R version 3.5.3## rgdal: version: 1.4-3, (SVN revision 828)
## Geospatial Data Abstraction Library extensions to R successfully loaded
## Loaded GDAL runtime: GDAL 2.2.3, released 2017/11/20
## Path to GDAL shared files: C:/Program Files/R/R-3.5.2/library/rgdal/gdal
## GDAL binary built with GEOS: TRUE
## Loaded PROJ.4 runtime: Rel. 4.9.3, 15 August 2016, [PJ_VERSION: 493]
## Path to PROJ.4 shared files: C:/Program Files/R/R-3.5.2/library/rgdal/proj
## Linking to sp version: 1.3-1library(raster)## Warning: package 'raster' was built under R version 3.5.3library(randomForest)## Warning: package 'randomForest' was built under R version 3.5.3## randomForest 4.6-14## Type rfNews() to see new features/changes/bug fixes.构建环境因子图层
setwd("E:/soilPaper/modling/EnFactor")
files<-c("DEM.tif","Slope.tif","PlanC.tif","ProfileC.tif","TWI.tif","Bedrock.tif","LSF.tif","VD.tif","RSP.tif")
covStack <- raster(files[1])
for (i in 2:length(files)) {
covStack <- stack(covStack, files[i])
}
#设置读取样点数据路径
point_shp_folder="E:/soilPaper/modling/allTrainingSamples_shp/"
#设置最终结果tif的保存路径
result_tif_folder="E:/soilPaper/modling/result_mapTif/"
##设置随机数种子
set.seed(2019)
soil_type<-c(1,3,7,20,21,23,30,31,501,502,503) # 41样本数量太少，删掉。
sampleNum_dict<-list()单因子分析
DEM -ok
ns3
SNA DEM ns3 15
mn="All_pSNA_DEM_NS3_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_DEM_NS3_pen15.shp", layer: "All_pSNA_DEM_NS3_pen15"
## with 1045 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 1040 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 16.73%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 3 0 0 0 1 0 0 0 0 0 0 0.25000000
## 20 0 124 10 0 0 0 0 0 0 17 9 0.22500000
## 21 0 21 36 0 0 0 0 0 0 3 0 0.40000000
## 23 0 0 0 80 0 6 1 0 0 12 3 0.21568627
## 3 0 0 0 0 50 0 0 5 0 0 0 0.09090909
## 30 0 0 0 16 0 15 2 0 0 0 0 0.54545455
## 31 0 0 0 2 0 1 37 0 0 0 0 0.07500000
## 501 0 0 0 0 7 0 0 165 0 0 0 0.04069767
## 502 0 2 2 0 0 0 0 7 10 0 1 0.54545455
## 503 0 12 0 4 0 0 0 0 0 223 7 0.09349593
## 7 0 10 0 0 0 0 0 0 2 11 123 0.15753425map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA DEM ns3 30
mn="All_pSNA_DEM_NS3_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_DEM_NS3_pen30.shp", layer: "All_pSNA_DEM_NS3_pen30"
## with 2045 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 2040 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 15.59%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 8 0 0 0 2 0 0 0 0 0 0 0.20000000
## 20 0 239 19 0 0 0 0 1 1 31 22 0.23642173
## 21 0 35 73 0 0 0 0 0 1 4 0 0.35398230
## 23 0 0 0 155 0 17 2 0 0 16 4 0.20103093
## 3 1 0 0 0 98 0 0 11 0 0 0 0.10909091
## 30 0 0 0 19 0 42 2 0 0 0 0 0.33333333
## 31 0 0 0 5 0 1 73 0 0 0 0 0.07594937
## 501 0 0 0 0 13 0 0 328 0 0 1 0.04093567
## 502 0 4 3 0 0 0 0 7 23 1 3 0.43902439
## 503 0 25 0 12 0 0 0 0 0 434 17 0.11065574
## 7 0 13 0 2 0 1 0 1 2 19 249 0.13240418map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA DEM ns3 45
mn="All_pSNA_DEM_NS3_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_DEM_NS3_pen45.shp", layer: "All_pSNA_DEM_NS3_pen45"
## with 3041 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 3036 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 14.49%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 15 0 0 0 1 0 0 0 0 0 0 0.06250000
## 20 0 368 27 0 0 0 0 0 2 32 35 0.20689655
## 21 0 54 106 0 0 0 0 0 1 6 0 0.36526946
## 23 0 0 0 241 0 19 6 0 0 17 4 0.16027875
## 3 2 0 0 0 139 0 0 23 0 0 0 0.15243902
## 30 0 0 0 16 0 71 4 0 0 0 1 0.22826087
## 31 0 0 0 7 0 5 106 0 0 0 0 0.10169492
## 501 0 0 0 0 20 0 0 490 1 0 2 0.04483431
## 502 0 0 2 0 0 0 0 7 45 2 5 0.26229508
## 503 0 40 1 15 0 0 0 0 0 649 22 0.10729023
## 7 0 25 0 4 0 1 0 3 3 25 366 0.14285714map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN DEM ns3 15
mn="All_pRAN_DEM_NS3_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_DEM_NS3_pen15.shp", layer: "All_pRAN_DEM_NS3_pen15"
## with 1045 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 1042 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 24.66%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 4 0 0 0 2 0 0 0 0 0 0 0.33333333
## 20 0 97 16 0 0 0 0 2 0 28 17 0.39375000
## 21 0 21 35 0 0 0 0 0 1 2 1 0.41666667
## 23 0 0 0 72 0 5 4 0 0 20 1 0.29411765
## 3 0 0 0 0 48 0 0 7 0 0 0 0.12727273
## 30 0 0 0 4 0 21 5 0 0 0 3 0.36363636
## 31 0 0 0 7 0 3 28 0 0 2 0 0.30000000
## 501 0 1 0 0 8 0 0 158 0 0 5 0.08139535
## 502 0 9 0 0 0 0 0 3 10 0 0 0.54545455
## 503 0 24 1 14 0 0 0 1 0 200 6 0.18699187
## 7 0 14 0 4 0 2 0 6 1 7 112 0.23287671map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN DEM ns3 30
mn="All_pRAN_DEM_NS3_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_DEM_NS3_pen30.shp", layer: "All_pRAN_DEM_NS3_pen30"
## with 2045 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 2042 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 21.55%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 12 0 0 0 0 0 0 0 0 0 0 0.0000000
## 20 0 211 22 0 0 0 0 4 1 48 27 0.3258786
## 21 0 34 69 0 0 0 0 1 1 8 0 0.3893805
## 23 0 0 0 139 0 7 10 0 0 32 6 0.2835052
## 3 1 0 0 0 86 0 0 23 0 0 0 0.2181818
## 30 0 0 0 15 0 39 5 0 0 0 4 0.3809524
## 31 0 0 0 12 0 2 65 0 0 0 0 0.1772152
## 501 0 2 0 0 23 0 0 305 1 0 11 0.1081871
## 502 0 2 2 0 0 0 0 1 29 0 7 0.2926829
## 503 0 31 7 22 0 0 0 0 2 414 12 0.1516393
## 7 0 20 0 3 0 2 0 12 2 15 233 0.1881533map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN DEM ns3 45
mn="All_pRAN_DEM_NS3_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_DEM_NS3_pen45.shp", layer: "All_pRAN_DEM_NS3_pen45"
## with 3041 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 3038 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 18.24%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 15 0 0 0 3 0 0 0 0 0 0 0.16666667
## 20 0 334 30 0 0 0 0 4 1 68 27 0.28017241
## 21 0 54 99 0 0 0 0 1 1 11 1 0.40718563
## 23 0 0 0 219 0 17 6 0 0 36 9 0.23693380
## 3 1 0 0 0 124 0 0 39 0 0 0 0.24390244
## 30 0 0 0 11 0 74 4 0 0 1 2 0.19565217
## 31 0 0 0 15 0 5 98 0 0 0 0 0.16949153
## 501 0 4 0 0 24 0 0 476 0 0 9 0.07212476
## 502 0 4 2 0 0 0 0 2 50 0 3 0.18032787
## 503 0 43 1 23 0 1 1 0 1 642 15 0.11691884
## 7 0 28 1 4 0 4 0 16 2 19 353 0.17330211map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")ns5
SNA DEM ns5 15
mn="All_pSNA_DEM_NS5_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_DEM_NS5_pen15.shp", layer: "All_pSNA_DEM_NS5_pen15"
## with 356 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 354 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 20.34%
## Confusion matrix:
## 20 21 23 3 30 31 501 502 503 7 class.error
## 20 30 5 0 0 0 0 0 0 7 6 0.37500000
## 21 10 9 0 0 0 0 0 1 2 0 0.59090909
## 23 0 0 26 0 1 1 0 0 0 2 0.13333333
## 3 0 0 0 19 0 0 1 0 0 0 0.05000000
## 30 0 0 5 0 2 0 0 0 0 1 0.75000000
## 31 0 0 1 0 0 12 0 0 0 0 0.07692308
## 501 0 0 0 1 0 0 65 0 0 0 0.01515152
## 502 2 2 0 0 0 0 3 1 0 0 0.87500000
## 503 1 1 1 0 0 0 0 0 77 6 0.10465116
## 7 3 0 0 0 0 0 0 0 9 41 0.22641509map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA DEM ns5 30
mn="All_pSNA_DEM_NS5_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_DEM_NS5_pen30.shp", layer: "All_pSNA_DEM_NS5_pen30"
## with 679 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 677 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 16.69%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 2 0 0 0 0 0 0 1.00000000
## 20 0 64 4 0 0 0 0 0 0 10 12 0.28888889
## 21 0 12 18 0 0 0 0 0 1 2 0 0.45454545
## 23 0 0 0 48 0 2 1 0 0 1 3 0.12727273
## 3 0 0 0 0 38 0 0 2 0 0 0 0.05000000
## 30 0 0 0 13 0 2 0 0 0 0 0 0.86666667
## 31 0 0 0 1 0 0 25 0 0 0 0 0.03846154
## 501 0 0 0 0 3 0 0 128 0 0 0 0.02290076
## 502 0 3 2 0 0 0 0 5 4 0 1 0.73333333
## 503 0 5 0 2 0 0 0 0 0 154 5 0.07228916
## 7 0 6 0 1 0 0 0 2 0 12 83 0.20192308map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA DEM ns5 45
mn="All_pSNA_DEM_NS5_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_DEM_NS5_pen45.shp", layer: "All_pSNA_DEM_NS5_pen45"
## with 995 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 993 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 16.21%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 3 0 0 0 1 0 0 0 0 0 0 0.25000000
## 20 0 94 8 0 0 0 0 0 1 12 13 0.26562500
## 21 0 15 29 0 0 0 0 0 0 1 0 0.35555556
## 23 0 0 0 71 0 7 1 0 0 2 2 0.14457831
## 3 0 0 0 0 54 0 0 5 0 0 0 0.08474576
## 30 0 0 0 13 0 6 0 0 0 0 0 0.68421053
## 31 0 0 0 1 0 0 38 0 0 0 0 0.02564103
## 501 0 0 0 0 7 0 0 190 0 0 0 0.03553299
## 502 0 5 2 0 0 0 0 8 6 0 1 0.72727273
## 503 0 11 0 3 0 0 0 0 0 223 11 0.10080645
## 7 0 11 0 1 0 0 0 2 0 17 118 0.20805369map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN DEM ns5 15
mn="All_pRAN_DEM_NS5_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_DEM_NS5_pen15.shp", layer: "All_pRAN_DEM_NS5_pen15"
## with 356 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 356 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 25%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 2 0 0 0 0 0 0 1.0000000
## 20 0 29 3 0 0 0 0 1 1 9 5 0.3958333
## 21 0 10 10 0 0 0 0 0 0 2 0 0.5454545
## 23 0 0 0 22 0 3 1 0 0 3 1 0.2666667
## 3 0 0 0 0 15 0 0 5 0 0 0 0.2500000
## 30 0 0 0 3 0 4 0 0 0 0 1 0.5000000
## 31 0 0 0 2 0 0 11 0 0 0 0 0.1538462
## 501 0 0 0 0 5 0 0 59 1 0 1 0.1060606
## 502 0 4 0 0 0 0 0 3 1 0 0 0.8750000
## 503 0 4 0 4 0 0 0 0 0 77 1 0.1046512
## 7 0 6 0 0 0 1 0 3 0 4 39 0.2641509map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN DEM ns5 30
mn="All_pRAN_DEM_NS5_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_DEM_NS5_pen30.shp", layer: "All_pRAN_DEM_NS5_pen30"
## with 679 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 679 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 19.73%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 2 0 0 0 2 0 0 0 0 0 0 0.50000000
## 20 0 57 10 0 0 0 0 1 2 10 10 0.36666667
## 21 0 13 19 0 0 0 0 0 0 1 0 0.42424242
## 23 0 0 0 40 0 4 2 0 0 6 3 0.27272727
## 3 1 0 0 0 34 0 0 5 0 0 0 0.15000000
## 30 0 0 0 8 0 4 3 0 0 0 0 0.73333333
## 31 0 0 0 6 0 1 19 0 0 0 0 0.26923077
## 501 0 1 0 0 5 0 0 124 0 0 1 0.05343511
## 502 0 3 0 0 0 0 0 1 9 1 1 0.40000000
## 503 0 6 0 3 0 0 0 0 0 153 4 0.07831325
## 7 0 8 0 2 0 0 0 4 0 6 84 0.19230769map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN DEM ns5 45
mn="All_pRAN_DEM_NS5_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_DEM_NS5_pen45.shp", layer: "All_pRAN_DEM_NS5_pen45"
## with 995 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 995 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 18.09%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 4 0 0 0 2 0 0 0 0 0 0 0.33333333
## 20 0 87 9 0 0 0 0 1 2 17 12 0.32031250
## 21 0 14 27 0 0 0 0 0 2 2 0 0.40000000
## 23 0 0 0 59 0 7 1 0 0 12 4 0.28915663
## 3 0 0 0 0 46 0 0 13 0 0 0 0.22033898
## 30 0 0 0 7 0 8 2 0 0 0 2 0.57894737
## 31 0 0 0 2 0 0 37 0 0 0 0 0.05128205
## 501 0 2 0 0 7 0 0 187 1 0 0 0.05076142
## 502 0 6 0 0 0 0 0 1 14 0 1 0.36363636
## 503 0 13 1 8 0 0 0 0 0 220 6 0.11290323
## 7 0 9 0 2 0 3 0 3 1 5 126 0.15436242map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")ns7
SNA DEM ns7 15
mn="All_pSNA_DEM_NS7_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_DEM_NS7_pen15.shp", layer: "All_pSNA_DEM_NS7_pen15"
## with 131 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 130 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 26.15%
## Confusion matrix:
## 20 21 23 3 30 31 501 502 503 7 class.error
## 20 12 1 0 0 0 0 0 0 3 2 0.3333333
## 21 2 1 0 0 0 0 0 1 0 0 0.7500000
## 23 0 0 10 0 1 0 0 0 0 1 0.1666667
## 3 0 0 0 6 0 0 1 0 0 0 0.1428571
## 30 0 0 2 0 0 0 0 0 1 0 1.0000000
## 31 0 0 1 0 0 3 0 0 0 0 0.2500000
## 501 1 0 0 0 0 0 24 0 0 0 0.0400000
## 502 2 0 0 0 0 0 2 0 0 0 1.0000000
## 503 4 0 0 0 0 0 0 0 24 3 0.2258065
## 7 1 0 0 0 0 0 0 0 5 16 0.2727273map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA DEM ns7 30
mn="All_pSNA_DEM_NS7_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_DEM_NS7_pen30.shp", layer: "All_pSNA_DEM_NS7_pen30"
## with 227 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 226 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 18.14%
## Confusion matrix:
## 20 21 23 3 30 31 501 502 503 7 class.error
## 20 16 1 0 0 0 0 0 0 7 3 0.4074074
## 21 3 3 0 0 0 0 0 0 0 0 0.5000000
## 23 0 0 18 0 2 0 0 0 0 0 0.1000000
## 3 0 0 0 13 0 0 0 0 0 0 0.0000000
## 30 0 0 3 0 0 0 0 0 0 1 1.0000000
## 31 0 0 0 0 0 7 0 0 0 0 0.0000000
## 501 0 0 0 0 0 0 49 0 0 0 0.0000000
## 502 1 1 0 0 0 0 3 0 0 1 1.0000000
## 503 3 0 0 0 0 0 0 0 49 4 0.1250000
## 7 1 0 0 0 0 0 0 0 7 30 0.2105263map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA DEM ns7 45
mn="All_pSNA_DEM_NS7_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_DEM_NS7_pen45.shp", layer: "All_pSNA_DEM_NS7_pen45"
## with 327 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 326 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 15.64%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.00000000
## 20 0 26 2 0 0 0 0 0 0 5 3 0.27777778
## 21 0 3 4 0 0 0 0 0 1 0 0 0.50000000
## 23 0 0 0 24 0 2 1 0 0 1 1 0.17241379
## 3 0 0 0 0 19 0 0 1 0 0 0 0.05000000
## 30 0 0 0 4 0 0 0 0 0 0 1 1.00000000
## 31 0 0 0 0 0 0 10 0 0 0 0 0.00000000
## 501 0 0 0 0 1 0 0 73 0 0 0 0.01351351
## 502 0 2 1 0 0 0 0 4 1 0 0 0.87500000
## 503 0 3 0 2 0 0 0 0 0 74 4 0.10843373
## 7 0 1 0 1 0 0 0 0 0 6 44 0.15384615map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN DEM ns7 15
mn="All_pRAN_DEM_NS7_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_DEM_NS7_pen15.shp", layer: "All_pRAN_DEM_NS7_pen15"
## with 131 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 131 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 29.77%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.0000000
## 20 0 9 1 0 0 0 0 0 0 2 6 0.5000000
## 21 0 2 2 0 0 0 0 0 0 0 0 0.5000000
## 23 0 0 0 7 0 1 0 0 0 4 0 0.4166667
## 3 1 0 0 0 5 0 0 1 0 0 0 0.2857143
## 30 0 0 0 3 0 0 0 0 0 0 0 1.0000000
## 31 0 0 0 0 0 0 4 0 0 0 0 0.0000000
## 501 0 1 0 0 1 0 0 22 0 0 1 0.1200000
## 502 0 1 0 0 0 0 0 1 1 0 1 0.7500000
## 503 0 3 0 3 0 0 0 0 0 24 1 0.2258065
## 7 0 2 0 0 0 0 0 0 0 2 18 0.1818182map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN DEM ns7 30
mn="All_pRAN_DEM_NS7_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_DEM_NS7_pen30.shp", layer: "All_pRAN_DEM_NS7_pen30"
## with 227 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 227 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 19.82%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.00000000
## 20 0 16 1 0 0 0 0 0 0 4 6 0.40740741
## 21 0 2 4 0 0 0 0 0 0 0 0 0.33333333
## 23 0 0 0 16 0 0 1 0 0 2 1 0.20000000
## 3 0 0 0 0 11 0 0 2 0 0 0 0.15384615
## 30 0 0 0 1 0 3 0 0 0 0 0 0.25000000
## 31 0 0 0 1 0 0 6 0 0 0 0 0.14285714
## 501 0 0 0 0 2 0 0 46 1 0 0 0.06122449
## 502 0 1 0 0 0 0 0 2 2 0 1 0.66666667
## 503 0 3 0 3 0 0 0 0 0 50 0 0.10714286
## 7 0 5 0 1 0 0 0 1 0 3 28 0.26315789map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN DEM ns7 45
mn="All_pRAN_DEM_NS7_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_DEM_NS7_pen45.shp", layer: "All_pRAN_DEM_NS7_pen45"
## with 327 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 327 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 16.51%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 1 0 0 0 1 0 0 0 0 0 0 0.50000000
## 20 0 25 1 0 0 0 0 0 0 5 5 0.30555556
## 21 0 3 4 0 0 0 0 0 1 0 0 0.50000000
## 23 0 0 0 22 0 2 1 0 0 4 0 0.24137931
## 3 0 0 0 0 18 0 0 2 0 0 0 0.10000000
## 30 0 0 0 2 0 3 0 0 0 0 0 0.40000000
## 31 0 0 0 1 0 0 9 0 0 0 0 0.10000000
## 501 0 0 0 0 2 0 0 71 0 0 1 0.04054054
## 502 0 2 0 0 0 0 0 2 3 0 1 0.62500000
## 503 0 2 0 4 0 0 0 0 0 75 2 0.09638554
## 7 0 4 0 0 0 0 0 1 0 5 42 0.19230769map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")Slope -ok
ns3
SNA Slope ns3 15
mn="All_pSNA_Slope_NS3_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_Slope_NS3_pen15.shp", layer: "All_pSNA_Slope_NS3_pen15"
## with 998 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 994 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 21.93%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 1 0 0 0 3 0 0 0 0 0 0 0.75000000
## 20 0 104 12 0 0 0 0 2 1 17 19 0.32903226
## 21 0 23 33 0 0 0 0 0 1 2 0 0.44067797
## 23 0 0 0 80 0 4 5 0 0 11 0 0.20000000
## 3 1 0 0 0 40 0 0 11 0 0 0 0.23076923
## 30 0 1 0 8 0 22 1 0 0 0 1 0.33333333
## 31 0 0 0 4 0 0 37 0 0 0 0 0.09756098
## 501 0 2 0 0 9 0 0 145 1 0 2 0.08805031
## 502 0 5 0 0 0 0 0 0 15 0 2 0.31818182
## 503 0 16 0 13 0 0 0 0 0 192 7 0.15789474
## 7 0 17 0 0 0 3 0 6 1 7 107 0.24113475map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA Slope ns3 30
mn="All_pSNA_Slope_NS3_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_Slope_NS3_pen30.shp", layer: "All_pSNA_Slope_NS3_pen30"
## with 1951 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 1947 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 19.26%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 7 0 0 0 3 0 0 0 0 0 0 0.30000000
## 20 0 213 13 0 0 0 0 4 2 42 29 0.29702970
## 21 0 41 63 0 0 0 0 0 2 5 0 0.43243243
## 23 0 0 0 154 0 8 5 0 0 21 3 0.19371728
## 3 2 0 0 0 84 0 0 18 0 0 0 0.19230769
## 30 0 1 0 7 0 51 1 0 0 0 3 0.19047619
## 31 0 0 0 2 0 1 79 0 0 0 0 0.03658537
## 501 0 1 0 0 13 0 0 288 0 2 13 0.09148265
## 502 0 7 2 0 0 0 0 0 28 1 3 0.31707317
## 503 0 29 2 17 0 1 0 0 2 388 10 0.13585746
## 7 0 24 1 0 0 1 0 14 1 18 217 0.21376812map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA Slope ns3 45
mn="All_pSNA_Slope_NS3_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_Slope_NS3_pen45.shp", layer: "All_pSNA_Slope_NS3_pen45"
## with 2889 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 2885 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 18.72%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 12 0 0 0 4 0 0 0 0 0 0 0.25000000
## 20 0 305 25 0 0 0 0 5 3 63 46 0.31767338
## 21 0 60 91 0 0 0 0 1 4 5 0 0.43478261
## 23 0 0 0 228 0 12 4 0 0 32 6 0.19148936
## 3 3 0 0 0 128 0 0 24 0 0 0 0.17419355
## 30 0 1 0 11 0 72 4 0 0 0 3 0.20879121
## 31 0 0 0 8 0 2 113 0 0 0 0 0.08130081
## 501 0 0 0 0 14 0 0 446 0 0 14 0.05907173
## 502 0 4 2 0 0 0 0 1 47 2 4 0.21666667
## 503 0 55 4 24 0 1 0 0 2 571 11 0.14520958
## 7 0 32 0 0 0 4 0 18 1 21 332 0.18627451map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN Slope ns3 15
mn="All_pRAN_Slope_NS3_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_Slope_NS3_pen15.shp", layer: "All_pRAN_Slope_NS3_pen15"
## with 998 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 996 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 27.41%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 6 0 0 0 0 0 0 0 0 0 0 0.00000000
## 20 0 104 14 0 0 0 0 3 0 22 12 0.32903226
## 21 0 33 20 0 0 0 0 0 0 5 1 0.66101695
## 23 0 0 0 69 0 9 3 0 0 16 3 0.31000000
## 3 0 0 0 0 41 0 0 11 0 0 0 0.21153846
## 30 0 0 0 8 0 17 4 0 0 2 2 0.48484848
## 31 0 0 0 9 0 4 28 0 0 0 0 0.31707317
## 501 0 2 0 0 8 0 0 144 0 0 5 0.09433962
## 502 0 6 0 0 0 0 0 3 12 0 1 0.45454545
## 503 0 19 4 15 0 1 0 0 0 177 12 0.22368421
## 7 0 15 0 1 0 4 0 7 1 8 105 0.25531915map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN Slope ns3 30
mn="All_pRAN_Slope_NS3_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_Slope_NS3_pen30.shp", layer: "All_pRAN_Slope_NS3_pen30"
## with 1951 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 1948 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 22.07%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 11 0 0 0 0 0 0 0 0 0 0 0.00000000
## 20 0 199 24 0 0 0 0 4 1 44 31 0.34323432
## 21 0 39 58 0 0 0 0 0 2 12 0 0.47747748
## 23 0 0 0 150 0 5 8 0 0 21 7 0.21465969
## 3 1 0 0 0 78 0 0 25 0 0 0 0.25000000
## 30 0 0 0 20 0 39 2 0 0 0 2 0.38095238
## 31 0 0 0 11 0 4 67 0 0 0 0 0.18292683
## 501 0 4 0 0 14 0 0 288 1 0 10 0.09148265
## 502 0 3 1 0 0 0 0 2 30 0 5 0.26829268
## 503 0 34 4 19 0 2 0 0 0 375 15 0.16481069
## 7 0 27 0 1 0 1 0 11 3 10 223 0.19202899map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN Slope ns3 45
mn="All_pRAN_Slope_NS3_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_Slope_NS3_pen45.shp", layer: "All_pRAN_Slope_NS3_pen45"
## with 2889 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 2886 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 19.4%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 15 0 0 0 2 0 0 0 0 0 0 0.1176471
## 20 0 334 21 1 0 1 0 5 2 60 23 0.2527964
## 21 0 52 91 0 0 0 0 0 2 14 2 0.4347826
## 23 0 0 0 220 0 16 7 0 0 31 8 0.2198582
## 3 2 0 0 0 121 0 0 32 0 0 0 0.2193548
## 30 0 0 0 19 0 63 6 0 0 0 3 0.3076923
## 31 0 0 0 13 0 6 104 0 0 0 0 0.1544715
## 501 0 5 0 0 28 0 0 430 2 0 9 0.0928270
## 502 0 5 2 0 0 0 0 3 44 0 6 0.2666667
## 503 0 44 4 27 0 1 1 0 1 573 17 0.1422156
## 7 0 31 0 2 0 3 0 18 6 17 331 0.1887255map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")ns5
SNA Slope ns5 15
mn="All_pSNA_Slope_NS5_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_Slope_NS5_pen15.shp", layer: "All_pSNA_Slope_NS5_pen15"
## with 336 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 336 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 24.4%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 2 0 0 0 0 0 0 1.00000000
## 20 0 29 4 0 0 0 0 1 0 6 6 0.36956522
## 21 0 6 12 0 0 0 0 1 0 2 0 0.42857143
## 23 0 0 0 19 0 3 0 0 0 5 2 0.34482759
## 3 0 0 0 0 16 0 0 3 0 0 0 0.15789474
## 30 0 0 0 4 0 4 1 0 0 0 0 0.55555556
## 31 0 0 0 1 0 0 12 0 0 0 0 0.07692308
## 501 0 1 0 0 4 0 0 55 0 0 1 0.09836066
## 502 0 3 0 0 0 0 0 0 3 1 0 0.57142857
## 503 0 5 0 3 0 0 0 0 0 67 3 0.14102564
## 7 0 5 1 0 0 1 0 3 0 4 37 0.27450980map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA Slope ns5 30
mn="All_pSNA_Slope_NS5_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_Slope_NS5_pen30.shp", layer: "All_pSNA_Slope_NS5_pen30"
## with 636 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 636 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 19.65%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 3 0 0 0 1 0 0 0 0 0 0 0.25000000
## 20 0 55 6 0 0 0 0 2 0 12 11 0.36046512
## 21 0 10 21 0 0 0 0 1 0 1 0 0.36363636
## 23 0 0 0 41 0 4 1 0 0 7 1 0.24074074
## 3 1 0 0 0 29 0 0 7 0 0 0 0.21621622
## 30 0 0 0 4 0 11 1 0 0 0 0 0.31250000
## 31 0 0 0 1 0 0 24 0 0 0 0 0.04000000
## 501 0 1 0 0 7 0 0 110 0 0 3 0.09090909
## 502 0 2 1 0 0 0 0 0 9 0 2 0.35714286
## 503 0 9 0 4 0 0 0 0 0 135 3 0.10596026
## 7 0 11 0 0 0 1 0 4 0 6 73 0.23157895map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA Slope ns5 45
mn="All_pSNA_Slope_NS5_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_Slope_NS5_pen45.shp", layer: "All_pSNA_Slope_NS5_pen45"
## with 931 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 929 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 18.3%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 2 0 0 0 2 0 0 0 0 0 0 0.50000000
## 20 0 75 9 0 0 0 0 2 0 16 18 0.37500000
## 21 0 12 31 0 0 0 0 1 1 1 1 0.34042553
## 23 0 0 0 62 0 5 2 0 0 9 2 0.22500000
## 3 0 0 0 0 46 0 0 9 0 0 0 0.16363636
## 30 0 0 0 5 0 16 1 0 0 0 0 0.27272727
## 31 0 0 0 1 0 0 36 0 0 0 0 0.02702703
## 501 0 1 0 0 5 0 0 170 0 0 5 0.06077348
## 502 0 3 1 0 0 0 0 0 13 0 3 0.35000000
## 503 0 13 0 6 0 0 0 0 0 199 5 0.10762332
## 7 0 15 0 2 0 1 0 5 0 8 109 0.22142857map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN Slope ns5 15
mn="All_pRAN_Slope_NS5_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_Slope_NS5_pen15.shp", layer: "All_pRAN_Slope_NS5_pen15"
## with 336 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 336 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 23.21%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 1 0 0 0 1 0 0 0 0 0 0 0.50000000
## 20 0 27 5 0 0 0 0 1 0 7 6 0.41304348
## 21 0 7 12 0 0 0 0 0 0 2 0 0.42857143
## 23 0 0 0 22 0 3 0 0 0 3 1 0.24137931
## 3 0 0 0 0 16 0 0 3 0 0 0 0.15789474
## 30 0 0 0 2 0 1 2 0 0 0 4 0.88888889
## 31 0 0 0 1 0 0 12 0 0 0 0 0.07692308
## 501 0 0 0 0 2 0 0 59 0 0 0 0.03278689
## 502 0 3 0 0 0 0 0 3 0 0 1 1.00000000
## 503 0 3 0 2 0 0 0 0 0 70 3 0.10256410
## 7 0 5 0 0 0 3 0 2 0 3 38 0.25490196map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN Slope ns5 30
mn="All_pRAN_Slope_NS5_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_Slope_NS5_pen30.shp", layer: "All_pRAN_Slope_NS5_pen30"
## with 636 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 636 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 21.07%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 4 0 0 0 0 0 0 0 0 0 0 0.00000000
## 20 0 58 10 0 0 0 0 1 1 8 8 0.32558140
## 21 0 20 12 0 0 0 0 0 0 1 0 0.63636364
## 23 0 0 0 39 0 5 2 0 0 4 4 0.27777778
## 3 1 0 0 0 28 0 0 8 0 0 0 0.24324324
## 30 0 0 0 4 0 11 1 0 0 0 0 0.31250000
## 31 0 0 0 3 0 0 22 0 0 0 0 0.12000000
## 501 0 0 0 0 6 0 0 110 1 0 4 0.09090909
## 502 0 3 0 0 0 0 0 3 6 0 2 0.57142857
## 503 0 6 1 4 0 0 0 0 0 136 4 0.09933775
## 7 0 8 0 3 0 0 0 3 1 4 76 0.20000000map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN Slope ns5 45
mn="All_pRAN_Slope_NS5_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_Slope_NS5_pen45.shp", layer: "All_pRAN_Slope_NS5_pen45"
## with 931 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 931 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 18.69%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 3 0 0 0 3 0 0 0 0 0 0 0.50000000
## 20 0 83 6 0 0 0 0 0 3 16 12 0.30833333
## 21 0 17 27 0 0 0 0 0 1 2 0 0.42553191
## 23 0 0 0 53 0 8 3 0 0 14 2 0.33750000
## 3 1 0 0 0 48 0 0 6 0 0 0 0.12727273
## 30 0 0 0 6 0 13 0 0 0 0 3 0.40909091
## 31 0 0 0 5 0 1 31 0 0 0 0 0.16216216
## 501 0 0 0 0 8 0 0 171 0 0 2 0.05524862
## 502 0 4 0 0 0 0 0 4 9 0 3 0.55000000
## 503 0 6 1 9 0 0 0 0 0 201 6 0.09865471
## 7 0 6 0 1 0 1 0 3 1 10 118 0.15714286map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")ns7
SNA Slope ns7 15
mn="All_pSNA_Slope_NS7_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_Slope_NS7_pen15.shp", layer: "All_pSNA_Slope_NS7_pen15"
## with 121 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 121 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 28.93%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.00000000
## 20 0 11 1 0 0 0 0 0 0 2 4 0.38888889
## 21 0 4 0 0 0 0 0 0 0 0 0 1.00000000
## 23 0 0 0 7 0 0 0 0 0 3 0 0.30000000
## 3 0 0 0 0 5 0 0 2 0 0 0 0.28571429
## 30 0 0 0 1 0 0 0 0 0 0 1 1.00000000
## 31 0 0 0 0 0 0 4 0 0 0 0 0.00000000
## 501 0 0 0 0 1 0 0 22 0 0 0 0.04347826
## 502 0 1 0 0 0 0 0 0 1 0 1 0.66666667
## 503 0 1 0 1 0 0 0 0 0 24 3 0.17241379
## 7 0 4 0 0 0 0 0 1 0 3 12 0.40000000map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA Slope ns7 30
mn="All_pSNA_Slope_NS7_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_Slope_NS7_pen30.shp", layer: "All_pSNA_Slope_NS7_pen30"
## with 214 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 214 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 20.56%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.00000000
## 20 0 16 1 0 0 0 0 0 1 4 5 0.40740741
## 21 0 2 3 0 0 0 0 0 0 0 0 0.40000000
## 23 0 0 0 13 0 0 0 0 0 4 0 0.23529412
## 3 0 0 0 0 12 0 0 2 0 0 0 0.14285714
## 30 0 0 0 0 0 3 0 0 0 0 0 0.00000000
## 31 0 0 0 0 0 0 7 0 0 0 0 0.00000000
## 501 0 0 0 0 1 0 0 43 0 0 2 0.06521739
## 502 0 1 0 0 2 0 0 1 0 0 1 1.00000000
## 503 0 1 0 3 0 0 0 0 0 45 4 0.15094340
## 7 0 2 0 0 0 0 0 1 0 5 28 0.22222222map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA Slope ns7 45
mn="All_pSNA_Slope_NS7_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_Slope_NS7_pen45.shp", layer: "All_pSNA_Slope_NS7_pen45"
## with 307 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 307 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 15.64%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 2 0 0 0 0 0 0 1.00000000
## 20 0 24 1 0 0 0 0 0 1 4 6 0.33333333
## 21 0 2 6 0 0 0 0 0 0 0 0 0.25000000
## 23 0 0 0 20 0 0 1 0 0 4 1 0.23076923
## 3 0 0 0 0 19 0 0 1 0 0 0 0.05000000
## 30 0 0 0 0 0 4 0 0 0 0 0 0.00000000
## 31 0 0 0 1 0 0 9 0 0 0 0 0.10000000
## 501 0 0 0 0 0 0 0 68 0 0 1 0.01449275
## 502 0 1 0 0 1 0 0 1 3 0 1 0.57142857
## 503 0 1 0 4 0 0 0 0 0 67 4 0.11842105
## 7 0 3 0 1 0 0 0 1 0 5 39 0.20408163map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN Slope ns7 15
mn="All_pRAN_Slope_NS7_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_Slope_NS7_pen15.shp", layer: "All_pRAN_Slope_NS7_pen15"
## with 121 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 121 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 21.49%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.00000000
## 20 0 13 0 0 0 0 0 0 0 2 3 0.27777778
## 21 0 2 2 0 0 0 0 0 0 0 0 0.50000000
## 23 0 0 0 7 0 0 1 0 0 2 0 0.30000000
## 3 1 0 0 0 5 0 0 1 0 0 0 0.28571429
## 30 0 0 0 2 0 0 0 0 0 0 0 1.00000000
## 31 0 0 0 2 0 0 2 0 0 0 0 0.50000000
## 501 0 0 0 0 0 0 0 23 0 0 0 0.00000000
## 502 0 1 1 0 0 0 0 1 0 0 0 1.00000000
## 503 0 2 0 0 0 0 0 0 0 27 0 0.06896552
## 7 0 1 0 1 0 0 0 0 0 2 16 0.20000000map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN Slope ns7 30
mn="All_pRAN_Slope_NS7_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_Slope_NS7_pen30.shp", layer: "All_pRAN_Slope_NS7_pen30"
## with 214 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 214 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 17.76%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.00000000
## 20 0 20 2 0 0 0 0 0 0 2 3 0.25925926
## 21 0 2 2 0 0 0 0 0 1 0 0 0.60000000
## 23 0 0 0 13 0 1 1 0 0 2 0 0.23529412
## 3 0 0 0 0 12 0 0 2 0 0 0 0.14285714
## 30 0 0 0 3 0 0 0 0 0 0 0 1.00000000
## 31 0 0 0 1 0 0 6 0 0 0 0 0.14285714
## 501 0 0 0 0 1 0 0 43 0 0 2 0.06521739
## 502 0 2 0 0 0 0 0 2 1 0 0 0.80000000
## 503 0 2 0 3 0 0 0 0 0 48 0 0.09433962
## 7 0 1 0 1 0 0 0 1 0 2 31 0.13888889map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN Slope ns7 45
mn="All_pRAN_Slope_NS7_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_Slope_NS7_pen45.shp", layer: "All_pRAN_Slope_NS7_pen45"
## with 307 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 307 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 17.26%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 1 0 0 0 1 0 0 0 0 0 0 0.50000000
## 20 0 25 2 0 0 0 0 0 0 3 6 0.30555556
## 21 0 4 4 0 0 0 0 0 0 0 0 0.50000000
## 23 0 0 0 19 0 1 2 0 0 4 0 0.26923077
## 3 2 0 0 0 16 0 0 2 0 0 0 0.20000000
## 30 0 0 0 4 0 0 0 0 0 0 0 1.00000000
## 31 0 0 0 1 0 0 9 0 0 0 0 0.10000000
## 501 0 0 0 0 2 0 0 66 0 0 1 0.04347826
## 502 0 2 0 0 0 0 0 2 3 0 0 0.57142857
## 503 0 3 0 3 0 0 0 0 0 68 2 0.10526316
## 7 0 3 0 0 0 0 0 1 0 2 43 0.12244898map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")PlanC -ok
ns3
SNA PlanC ns3 15
mn="All_pSNA_PlanC_NS3_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_PlanC_NS3_pen15.shp", layer: "All_pSNA_PlanC_NS3_pen15"
## with 1013 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 1011 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 22.06%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 4 0 0 0 2 0 0 0 0 0 0 0.3333333
## 20 0 98 15 0 0 0 0 1 0 25 17 0.3717949
## 21 0 25 32 0 0 0 0 0 0 2 0 0.4576271
## 23 0 0 0 88 0 3 1 0 0 8 0 0.1200000
## 3 1 0 0 0 41 0 0 9 0 0 0 0.1960784
## 30 0 0 0 6 0 23 0 0 0 2 2 0.3030303
## 31 0 0 0 7 0 1 34 0 0 0 0 0.1904762
## 501 0 2 0 0 8 0 0 149 0 0 6 0.0969697
## 502 0 2 0 0 0 0 0 2 9 0 8 0.5714286
## 503 0 24 2 6 0 2 0 0 0 193 6 0.1716738
## 7 0 10 0 0 0 2 0 7 3 6 117 0.1931034map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA PlanC ns3 30
mn="All_pSNA_PlanC_NS3_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_PlanC_NS3_pen30.shp", layer: "All_pSNA_PlanC_NS3_pen30"
## with 1976 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 1973 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 19.97%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 8 0 0 0 2 0 0 0 0 0 0 0.20000000
## 20 0 196 28 0 0 0 0 3 1 46 29 0.35313531
## 21 0 41 57 0 0 0 0 2 0 11 1 0.49107143
## 23 0 0 0 156 0 5 3 0 0 24 1 0.17460317
## 3 2 0 0 0 80 0 0 19 0 0 0 0.20792079
## 30 0 0 0 9 0 49 1 0 0 1 3 0.22222222
## 31 0 0 0 5 0 1 78 0 0 0 0 0.07142857
## 501 0 1 1 0 12 0 0 306 1 0 9 0.07272727
## 502 0 5 1 0 0 0 0 2 24 1 7 0.40000000
## 503 0 43 1 14 0 3 0 0 0 390 9 0.15217391
## 7 0 17 0 0 0 1 0 13 2 13 235 0.16370107map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA PlanC ns3 45
mn="All_pSNA_PlanC_NS3_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_PlanC_NS3_pen45.shp", layer: "All_pSNA_PlanC_NS3_pen45"
## with 2940 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 2937 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 19.85%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 13 0 0 0 2 0 0 0 0 0 0 0.13333333
## 20 0 295 29 0 0 0 0 5 3 72 46 0.34444444
## 21 0 59 89 0 0 0 0 2 1 13 1 0.46060606
## 23 0 0 0 224 0 11 6 0 0 36 4 0.20284698
## 3 2 0 0 0 113 0 0 37 0 0 0 0.25657895
## 30 0 0 0 10 0 72 2 0 0 1 6 0.20879121
## 31 0 0 0 5 0 1 119 0 0 0 0 0.04800000
## 501 0 1 1 0 28 0 0 447 1 0 15 0.09330629
## 502 0 2 1 0 0 0 0 2 47 1 5 0.18965517
## 503 0 58 1 20 0 2 0 0 1 591 15 0.14098837
## 7 0 33 0 0 0 3 0 20 1 18 344 0.17899761map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN PlanC ns3 15
mn="All_pRAN_PlanC_NS3_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_PlanC_NS3_pen15.shp", layer: "All_pRAN_PlanC_NS3_pen15"
## with 1013 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 1011 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 25.72%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 5 0 0 0 1 0 0 0 0 0 0 0.1666667
## 20 0 92 19 0 0 0 0 1 0 28 16 0.4102564
## 21 0 19 36 0 0 0 0 0 0 4 0 0.3898305
## 23 0 0 0 73 0 7 3 0 0 14 3 0.2700000
## 3 1 0 0 0 37 0 0 13 0 0 0 0.2745098
## 30 0 0 0 9 0 17 4 0 0 1 2 0.4848485
## 31 0 0 0 6 0 3 33 0 0 0 0 0.2142857
## 501 0 2 0 0 12 0 0 145 0 0 6 0.1212121
## 502 0 5 2 0 0 0 0 3 10 1 0 0.5238095
## 503 0 18 2 10 0 1 0 0 2 192 8 0.1759657
## 7 0 13 0 1 0 0 0 9 1 10 111 0.2344828map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN PlanC ns3 30
mn="All_pRAN_PlanC_NS3_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_PlanC_NS3_pen30.shp", layer: "All_pRAN_PlanC_NS3_pen30"
## with 1976 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 1973 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 22.05%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 9 0 0 0 1 0 0 0 0 0 0 0.10000000
## 20 0 191 33 0 0 0 0 5 0 48 26 0.36963696
## 21 0 43 60 0 0 0 0 0 0 9 0 0.46428571
## 23 0 0 0 134 0 13 3 0 0 32 7 0.29100529
## 3 0 0 0 0 78 0 0 23 0 0 0 0.22772277
## 30 0 0 0 12 0 44 4 0 0 1 2 0.30158730
## 31 0 0 0 9 0 4 70 0 0 1 0 0.16666667
## 501 0 1 0 0 16 0 0 304 1 0 8 0.07878788
## 502 0 7 0 0 0 0 0 5 26 1 1 0.35000000
## 503 0 26 3 19 0 1 1 0 1 399 10 0.13260870
## 7 0 22 0 3 0 0 0 17 1 15 223 0.20640569map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN PlanC ns3 45
mn="All_pRAN_PlanC_NS3_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_PlanC_NS3_pen45.shp", layer: "All_pRAN_PlanC_NS3_pen45"
## with 2940 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 2937 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 18.05%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 14 0 0 0 1 0 0 0 0 0 0 0.06666667
## 20 0 320 32 1 0 0 0 2 3 63 29 0.28888889
## 21 0 46 106 0 0 0 0 0 2 11 0 0.35757576
## 23 0 0 0 225 0 14 4 0 0 29 9 0.19928826
## 3 2 0 0 0 113 0 0 37 0 0 0 0.25657895
## 30 0 0 0 13 0 70 4 0 0 2 2 0.23076923
## 31 0 0 0 12 0 4 108 0 0 1 0 0.13600000
## 501 0 4 0 0 25 0 0 455 1 0 8 0.07707911
## 502 0 2 2 0 0 0 0 2 50 1 1 0.13793103
## 503 0 45 4 27 0 1 1 0 2 592 16 0.13953488
## 7 0 20 0 4 0 3 0 16 2 20 354 0.15513126map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")ns5
SNA PlanC ns5 15
mn="All_pSNA_PlanC_NS5_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_PlanC_NS5_pen15.shp", layer: "All_pSNA_PlanC_NS5_pen15"
## with 345 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 345 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 26.96%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 2 0 0 0 0 0 0 1.0000000
## 20 0 22 6 0 0 0 0 0 0 9 8 0.5111111
## 21 0 8 12 0 0 0 0 0 0 2 0 0.4545455
## 23 0 0 0 21 0 2 1 0 0 5 0 0.2758621
## 3 1 0 0 0 13 0 0 4 0 0 0 0.2777778
## 30 0 0 0 2 0 7 0 0 0 0 0 0.2222222
## 31 0 0 0 2 0 0 12 0 0 0 0 0.1428571
## 501 0 0 0 0 7 0 0 56 0 0 1 0.1250000
## 502 0 2 0 0 0 0 0 2 1 1 1 0.8571429
## 503 0 7 1 3 0 0 0 0 0 68 3 0.1707317
## 7 0 6 0 0 0 1 0 1 0 5 40 0.2452830map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA PlanC ns5 30
mn="All_pSNA_PlanC_NS5_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_PlanC_NS5_pen30.shp", layer: "All_pSNA_PlanC_NS5_pen30"
## with 650 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 649 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 21.26%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 1 0 0 0 2 0 0 0 0 0 0 0.66666667
## 20 0 46 11 0 0 0 0 0 0 14 13 0.45238095
## 21 0 15 15 0 0 0 0 0 1 3 0 0.55882353
## 23 0 0 0 44 0 2 3 0 0 6 0 0.20000000
## 3 1 0 0 0 26 0 0 9 0 0 0 0.27777778
## 30 0 0 0 3 0 9 2 0 0 0 0 0.35714286
## 31 0 0 0 3 0 0 24 0 0 0 0 0.11111111
## 501 0 0 0 0 9 0 0 116 0 0 2 0.08661417
## 502 0 5 0 0 0 0 0 1 6 1 1 0.57142857
## 503 0 7 3 5 0 0 0 0 0 135 5 0.12903226
## 7 0 5 0 0 0 1 0 1 1 3 89 0.11000000map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA PlanC ns5 45
mn="All_pSNA_PlanC_NS5_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_PlanC_NS5_pen45.shp", layer: "All_pSNA_PlanC_NS5_pen45"
## with 954 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 953 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 19.52%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 2 0 0 0 2 0 0 0 0 0 0 0.50000000
## 20 0 71 11 0 0 0 0 0 0 22 16 0.40833333
## 21 0 17 23 0 0 0 0 0 1 4 0 0.48888889
## 23 0 0 0 63 0 3 5 0 0 9 0 0.21250000
## 3 2 0 0 0 35 0 0 17 0 0 0 0.35185185
## 30 0 0 0 2 0 17 2 0 0 0 0 0.19047619
## 31 0 0 0 4 0 0 36 0 0 0 0 0.10000000
## 501 0 0 0 0 14 0 0 173 0 0 3 0.08947368
## 502 0 2 0 0 0 0 0 2 12 1 3 0.40000000
## 503 0 9 1 7 0 1 0 0 0 207 7 0.10775862
## 7 0 7 0 0 0 1 0 5 0 6 128 0.12925170map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN PlanC ns5 15
mn="All_pRAN_PlanC_NS5_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_PlanC_NS5_pen15.shp", layer: "All_pRAN_PlanC_NS5_pen15"
## with 345 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 345 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 28.41%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 2 0 0 0 0 0 0 1.0000000
## 20 0 25 5 0 0 0 0 0 1 7 7 0.4444444
## 21 0 11 8 0 0 0 0 0 1 2 0 0.6363636
## 23 0 0 0 18 0 3 0 0 0 4 4 0.3793103
## 3 0 0 0 0 14 0 0 4 0 0 0 0.2222222
## 30 0 0 0 6 0 1 2 0 0 0 0 0.8888889
## 31 0 0 0 1 0 2 10 0 0 0 1 0.2857143
## 501 0 2 0 0 1 0 0 59 0 0 2 0.0781250
## 502 0 3 0 0 0 0 0 1 2 1 0 0.7142857
## 503 0 2 2 2 0 0 0 0 0 73 3 0.1097561
## 7 0 5 0 1 0 1 0 5 0 4 37 0.3018868map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN PlanC ns5 30
mn="All_pRAN_PlanC_NS5_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_PlanC_NS5_pen30.shp", layer: "All_pRAN_PlanC_NS5_pen30"
## with 650 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 650 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 19.38%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 4 0 0 0 0 0 0 0 0 0 0 0.00000000
## 20 0 55 10 0 0 0 0 1 0 11 7 0.34523810
## 21 0 15 16 0 0 0 0 0 1 1 1 0.52941176
## 23 0 0 0 42 0 5 1 0 0 5 2 0.23636364
## 3 1 0 0 0 33 0 0 2 0 0 0 0.08333333
## 30 0 0 0 7 0 6 1 0 0 0 0 0.57142857
## 31 0 0 0 4 0 0 23 0 0 0 0 0.14814815
## 501 0 2 0 0 7 0 0 116 1 0 1 0.08661417
## 502 0 4 1 0 0 0 0 2 5 0 2 0.64285714
## 503 0 3 2 1 0 0 0 0 0 140 9 0.09677419
## 7 0 6 0 0 0 0 0 3 1 6 84 0.16000000map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN PlanC ns5 45
mn="All_pRAN_PlanC_NS5_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_PlanC_NS5_pen45.shp", layer: "All_pRAN_PlanC_NS5_pen45"
## with 954 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 954 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 17.92%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 4 0 0 0 1 0 0 0 0 0 0 0.20000000
## 20 0 80 7 0 0 0 0 1 0 16 16 0.33333333
## 21 0 20 20 0 0 0 0 0 0 4 1 0.55555556
## 23 0 0 0 55 0 7 4 0 0 11 3 0.31250000
## 3 0 0 0 0 45 0 0 9 0 0 0 0.16666667
## 30 0 0 0 6 0 13 0 0 0 0 2 0.38095238
## 31 0 0 0 4 0 1 35 0 0 0 0 0.12500000
## 501 0 1 0 0 8 0 0 178 0 0 3 0.06315789
## 502 0 1 0 0 0 0 0 3 15 0 1 0.25000000
## 503 0 5 3 6 0 0 0 0 0 212 6 0.08620690
## 7 0 8 0 0 0 1 0 4 0 8 126 0.14285714map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")ns7
SNA PlanC ns7 15
mn="All_pSNA_PlanC_NS7_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_PlanC_NS7_pen15.shp", layer: "All_pSNA_PlanC_NS7_pen15"
## with 129 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 129 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 30.23%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.0000000
## 20 0 9 0 0 0 0 0 0 0 4 5 0.5000000
## 21 0 1 3 0 0 0 0 0 0 0 0 0.2500000
## 23 0 0 0 6 0 1 0 0 0 3 0 0.4000000
## 3 0 0 0 0 6 0 0 1 0 0 0 0.1428571
## 30 0 0 0 1 0 0 0 0 0 1 0 1.0000000
## 31 0 0 0 0 0 0 4 0 0 0 0 0.0000000
## 501 0 0 0 0 2 0 0 23 0 0 0 0.0800000
## 502 0 0 0 0 0 0 0 1 2 1 0 0.5000000
## 503 0 4 1 3 0 0 0 0 0 22 1 0.2903226
## 7 0 4 0 0 0 0 0 1 0 3 15 0.3478261map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA PlanC ns7 30
mn="All_pSNA_PlanC_NS7_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_PlanC_NS7_pen30.shp", layer: "All_pSNA_PlanC_NS7_pen30"
## with 227 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 227 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 22.03%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.00000000
## 20 0 17 1 0 0 0 0 0 0 3 7 0.39285714
## 21 0 2 3 0 0 0 0 0 0 0 0 0.40000000
## 23 0 0 0 15 0 0 1 0 0 3 0 0.21052632
## 3 0 0 0 0 9 0 0 4 0 0 0 0.30769231
## 30 0 0 0 1 0 1 0 0 0 1 0 0.66666667
## 31 0 0 0 1 0 0 6 0 0 0 0 0.14285714
## 501 0 0 0 0 3 0 0 46 0 0 0 0.06122449
## 502 0 1 0 0 0 0 0 2 3 0 0 0.50000000
## 503 0 5 0 2 0 0 0 0 0 48 2 0.15789474
## 7 0 5 0 0 0 0 0 1 0 4 29 0.25641026map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA PlanC ns7 45
mn="All_pSNA_PlanC_NS7_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_PlanC_NS7_pen45.shp", layer: "All_pSNA_PlanC_NS7_pen45"
## with 320 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 319 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 20.69%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.00000000
## 20 0 18 1 0 0 0 0 0 0 9 7 0.48571429
## 21 0 4 4 0 0 0 0 0 0 0 0 0.50000000
## 23 0 0 0 19 0 0 1 0 0 6 0 0.26923077
## 3 0 0 0 0 13 0 0 6 0 0 0 0.31578947
## 30 0 0 0 1 0 2 0 0 0 1 0 0.50000000
## 31 0 0 0 0 0 0 11 0 0 0 0 0.00000000
## 501 0 0 0 0 4 0 0 69 0 0 0 0.05479452
## 502 0 2 0 0 0 0 0 2 4 0 0 0.50000000
## 503 0 6 0 4 0 0 0 0 0 69 2 0.14814815
## 7 0 4 0 0 0 0 0 1 0 4 44 0.16981132map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN PlanC ns7 15
mn="All_pRAN_PlanC_NS7_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_PlanC_NS7_pen15.shp", layer: "All_pRAN_PlanC_NS7_pen15"
## with 129 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 129 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 25.58%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.0000000
## 20 0 9 2 0 0 0 0 0 0 1 6 0.5000000
## 21 0 3 0 0 0 0 0 0 1 0 0 1.0000000
## 23 0 0 0 7 0 0 1 0 0 2 0 0.3000000
## 3 0 0 0 0 5 0 0 2 0 0 0 0.2857143
## 30 0 0 0 0 0 2 0 0 0 0 0 0.0000000
## 31 0 0 0 0 0 0 4 0 0 0 0 0.0000000
## 501 0 0 0 0 1 0 0 23 1 0 0 0.0800000
## 502 0 1 0 0 0 0 0 1 2 0 0 0.5000000
## 503 0 2 0 2 0 0 0 0 0 26 1 0.1612903
## 7 0 3 0 0 0 0 0 0 0 2 18 0.2173913map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN PlanC ns7 30
mn="All_pRAN_PlanC_NS7_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_PlanC_NS7_pen30.shp", layer: "All_pRAN_PlanC_NS7_pen30"
## with 227 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 227 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 18.5%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.00000000
## 20 0 18 1 0 0 0 0 0 0 3 6 0.35714286
## 21 0 2 2 0 0 0 0 0 1 0 0 0.60000000
## 23 0 0 0 14 0 2 0 0 0 3 0 0.26315789
## 3 0 0 0 0 12 0 0 1 0 0 0 0.07692308
## 30 0 0 0 2 0 0 1 0 0 0 0 1.00000000
## 31 0 0 0 1 0 0 6 0 0 0 0 0.14285714
## 501 0 0 0 0 1 0 0 48 0 0 0 0.02040816
## 502 0 1 0 0 0 0 0 2 2 0 1 0.66666667
## 503 0 1 0 1 0 0 0 0 0 53 2 0.07017544
## 7 0 4 0 1 0 0 0 2 0 2 30 0.23076923map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN PlanC ns7 45
mn="All_pRAN_PlanC_NS7_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_PlanC_NS7_pen45.shp", layer: "All_pRAN_PlanC_NS7_pen45"
## with 320 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 320 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 17.81%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 2 0 0 0 0 0 0 0 0 0 0 0.00000000
## 20 0 26 0 0 0 0 0 0 0 3 6 0.25714286
## 21 0 1 6 0 0 0 0 0 1 0 0 0.25000000
## 23 0 0 0 18 0 2 1 0 0 5 0 0.30769231
## 3 0 0 0 0 15 0 0 4 0 0 0 0.21052632
## 30 0 0 0 4 0 0 0 0 0 0 0 1.00000000
## 31 0 0 0 1 0 0 10 0 0 0 0 0.09090909
## 501 0 0 0 0 2 0 0 69 1 0 1 0.05479452
## 502 0 2 0 0 0 0 0 2 3 0 1 0.62500000
## 503 0 4 0 5 0 0 0 0 0 72 0 0.11111111
## 7 0 5 0 1 0 0 0 1 0 4 42 0.20754717map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")ProfileC -ok
ns3
SNA ProfileC ns3 15
mn="All_pSNA_ProfileC_NS3_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_ProfileC_NS3_pen15.shp", layer: "All_pSNA_ProfileC_NS3_pen15"
## with 1000 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 996 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 24.5%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 2 0 0 0 1 0 0 0 0 0 0 0.33333333
## 20 0 84 21 0 0 0 0 1 1 31 18 0.46153846
## 21 0 23 31 0 0 0 0 0 0 3 0 0.45614035
## 23 0 0 0 69 0 4 2 0 0 17 5 0.28865979
## 3 1 0 0 0 43 0 0 9 0 0 0 0.18867925
## 30 0 0 0 8 0 19 0 0 0 1 5 0.42424242
## 31 0 0 0 2 0 1 36 0 0 0 0 0.07692308
## 501 0 0 0 0 7 0 0 150 0 0 4 0.06832298
## 502 0 6 0 0 0 0 0 2 11 1 1 0.47619048
## 503 0 25 2 9 0 1 0 0 0 191 5 0.18025751
## 7 0 8 0 1 0 2 0 9 0 7 116 0.18881119map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA ProfileC ns3 30
mn="All_pSNA_ProfileC_NS3_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_ProfileC_NS3_pen30.shp", layer: "All_pSNA_ProfileC_NS3_pen30"
## with 1950 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 1946 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 21.69%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 4 0 0 0 4 0 0 0 0 0 0 0.50000000
## 20 0 179 37 0 0 0 0 3 0 51 33 0.40924092
## 21 0 46 53 0 0 0 0 0 0 9 0 0.50925926
## 23 0 0 0 143 0 12 3 0 0 23 5 0.23118280
## 3 2 0 0 0 94 0 0 9 0 0 0 0.10476190
## 30 0 0 0 6 0 46 2 0 0 2 6 0.25806452
## 31 0 0 0 5 0 2 70 0 0 1 0 0.10256410
## 501 0 3 0 0 16 0 0 292 0 0 10 0.09034268
## 502 0 3 0 0 0 0 0 3 31 1 2 0.22500000
## 503 0 41 2 13 0 1 0 0 0 394 7 0.13973799
## 7 0 19 0 2 0 1 0 24 1 12 218 0.21299639map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA ProfileC ns3 45
mn="All_pSNA_ProfileC_NS3_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_ProfileC_NS3_pen45.shp", layer: "All_pSNA_ProfileC_NS3_pen45"
## with 2899 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 2895 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 19.45%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 10 0 0 0 3 0 0 0 0 0 0 0.23076923
## 20 0 288 48 0 0 0 0 7 0 70 36 0.35857461
## 21 0 56 89 0 0 0 0 0 0 14 1 0.44375000
## 23 0 0 0 228 0 11 3 0 0 28 4 0.16788321
## 3 4 0 0 0 140 0 0 13 0 0 0 0.10828025
## 30 0 0 0 8 0 71 2 0 0 4 6 0.21978022
## 31 0 0 0 5 0 6 105 0 0 1 0 0.10256410
## 501 0 4 0 0 25 0 0 436 0 0 16 0.09355509
## 502 0 1 1 0 0 0 0 3 46 4 2 0.19298246
## 503 0 57 3 17 0 4 0 0 2 590 12 0.13868613
## 7 0 28 1 3 0 2 0 33 1 14 329 0.19951338map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN ProfileC ns3 15
mn="All_pRAN_ProfileC_NS3_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_ProfileC_NS3_pen15.shp", layer: "All_pRAN_ProfileC_NS3_pen15"
## with 1000 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 998 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 25.75%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 4 0 0 0 1 0 0 0 0 0 0 0.2000000
## 20 0 94 19 1 0 0 0 1 1 28 12 0.3974359
## 21 0 30 25 0 0 0 0 0 0 2 0 0.5614035
## 23 0 0 0 64 0 10 3 0 0 17 3 0.3402062
## 3 0 0 0 0 39 0 0 14 0 0 0 0.2641509
## 30 0 0 0 7 0 20 5 0 0 0 1 0.3939394
## 31 0 0 0 6 0 4 29 0 0 0 0 0.2564103
## 501 0 1 0 0 10 0 0 141 2 0 7 0.1242236
## 502 0 3 0 0 0 0 0 3 13 0 2 0.3809524
## 503 0 19 2 6 0 0 0 0 1 200 5 0.1416309
## 7 0 12 0 1 0 2 0 6 1 9 112 0.2167832map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN ProfileC ns3 30
mn="All_pRAN_ProfileC_NS3_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_ProfileC_NS3_pen30.shp", layer: "All_pRAN_ProfileC_NS3_pen30"
## with 1950 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 1946 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 22.1%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 7 0 0 0 1 0 0 0 0 0 0 0.12500000
## 20 0 220 23 2 0 1 0 3 1 34 19 0.27392739
## 21 0 46 56 0 0 0 0 0 1 4 1 0.48148148
## 23 0 2 0 132 0 16 4 0 0 25 7 0.29032258
## 3 0 0 0 0 78 0 0 27 0 0 0 0.25714286
## 30 0 0 0 13 0 40 5 0 0 1 3 0.35483871
## 31 0 0 0 11 0 3 64 0 0 0 0 0.17948718
## 501 0 0 0 0 16 0 0 297 1 0 7 0.07476636
## 502 0 6 1 0 0 0 0 2 27 0 4 0.32500000
## 503 0 38 2 23 0 1 0 0 1 380 13 0.17030568
## 7 0 26 0 4 0 2 0 15 2 13 215 0.22382671map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN ProfileC ns3 45
mn="All_pRAN_ProfileC_NS3_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_ProfileC_NS3_pen45.shp", layer: "All_pRAN_ProfileC_NS3_pen45"
## with 2899 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 2894 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 19.38%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 11 0 0 0 1 0 0 0 0 0 0 0.08333333
## 20 0 329 23 1 0 0 0 6 2 61 27 0.26726058
## 21 0 56 93 0 0 0 0 1 0 8 2 0.41875000
## 23 0 0 0 202 0 24 4 0 0 37 7 0.26277372
## 3 1 0 0 0 121 0 0 35 0 0 0 0.22929936
## 30 0 0 0 20 0 64 5 0 0 0 2 0.29670330
## 31 0 0 0 13 0 3 100 0 0 1 0 0.14529915
## 501 0 6 1 0 24 0 0 438 1 0 11 0.08939709
## 502 0 3 1 0 0 0 0 2 44 1 6 0.22807018
## 503 0 44 1 23 0 2 0 0 1 597 17 0.12846715
## 7 0 27 0 1 0 5 0 24 1 19 334 0.18734793map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")ns5
SNA ProfileC ns5 15
mn="All_pSNA_ProfileC_NS5_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_ProfileC_NS5_pen15.shp", layer: "All_pSNA_ProfileC_NS5_pen15"
## with 340 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 338 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 26.33%
## Confusion matrix:
## 20 21 23 3 30 31 501 502 503 7 class.error
## 20 26 5 0 0 0 0 0 0 7 8 0.43478261
## 21 15 6 0 0 0 0 0 0 2 0 0.73913043
## 23 0 0 19 0 2 0 0 0 6 1 0.32142857
## 3 0 0 0 16 0 0 2 0 0 0 0.11111111
## 30 0 0 3 0 5 0 0 0 0 1 0.44444444
## 31 0 0 0 0 0 14 0 0 0 0 0.00000000
## 501 0 0 0 4 0 0 54 0 0 1 0.08474576
## 502 1 0 0 0 0 0 1 4 0 1 0.42857143
## 503 7 3 2 0 0 0 0 0 67 3 0.18292683
## 7 3 0 1 0 1 0 4 0 5 38 0.26923077map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA ProfileC ns5 30
mn="All_pSNA_ProfileC_NS5_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_ProfileC_NS5_pen30.shp", layer: "All_pSNA_ProfileC_NS5_pen30"
## with 638 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 636 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 21.54%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 2 0 0 0 0 0 0 1.00000000
## 20 0 47 9 0 0 0 0 0 0 16 12 0.44047619
## 21 0 14 19 0 0 0 0 0 0 2 0 0.45714286
## 23 0 0 0 43 0 2 2 0 0 5 1 0.18867925
## 3 1 0 0 0 30 0 0 4 0 0 0 0.14285714
## 30 0 0 0 3 0 12 0 0 0 0 0 0.20000000
## 31 0 0 0 1 0 0 26 0 0 0 0 0.03703704
## 501 0 0 0 0 6 0 0 108 1 0 2 0.07692308
## 502 0 2 0 0 0 0 0 3 7 0 2 0.50000000
## 503 0 12 2 3 0 0 0 0 0 135 6 0.14556962
## 7 0 4 0 0 0 2 0 9 0 9 72 0.25000000map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA ProfileC ns5 45
mn="All_pSNA_ProfileC_NS5_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_ProfileC_NS5_pen45.shp", layer: "All_pSNA_ProfileC_NS5_pen45"
## with 936 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 934 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 19.27%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 3 0 0 0 0 0 0 0 0 0 0 0.00000000
## 20 0 72 12 0 0 0 0 0 0 21 16 0.40495868
## 21 0 18 27 0 0 0 0 0 0 2 1 0.43750000
## 23 0 0 0 61 0 3 3 0 0 9 2 0.21794872
## 3 0 0 0 0 46 0 0 6 0 0 0 0.11538462
## 30 0 0 0 6 0 16 0 0 0 0 0 0.27272727
## 31 0 0 0 2 0 0 39 0 0 0 0 0.04878049
## 501 0 2 0 0 9 0 0 162 0 0 2 0.07428571
## 502 0 3 0 0 1 0 0 2 11 1 2 0.45000000
## 503 0 14 3 5 0 0 0 0 0 208 5 0.11489362
## 7 0 8 0 1 0 2 0 11 1 7 109 0.21582734map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN ProfileC ns5 15
mn="All_pRAN_ProfileC_NS5_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_ProfileC_NS5_pen15.shp", layer: "All_pRAN_ProfileC_NS5_pen15"
## with 340 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 340 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 22.65%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 2 0 0 0 0 0 0 1.00000000
## 20 0 25 7 0 0 0 0 0 1 8 5 0.45652174
## 21 0 6 16 0 0 0 0 0 0 1 0 0.30434783
## 23 0 0 0 20 0 3 1 0 0 3 1 0.28571429
## 3 0 0 0 0 15 0 0 3 0 0 0 0.16666667
## 30 0 0 0 3 0 5 0 0 0 0 1 0.44444444
## 31 0 0 0 3 0 0 11 0 0 0 0 0.21428571
## 501 0 0 0 0 5 0 0 52 1 0 1 0.11864407
## 502 0 2 0 0 0 0 0 2 2 0 1 0.71428571
## 503 0 4 0 2 0 1 0 0 0 74 1 0.09756098
## 7 0 3 0 0 0 0 0 1 0 5 43 0.17307692map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN ProfileC ns5 30
mn="All_pRAN_ProfileC_NS5_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_ProfileC_NS5_pen30.shp", layer: "All_pRAN_ProfileC_NS5_pen30"
## with 638 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 638 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 18.5%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 3 0 0 0 1 0 0 0 0 0 0 0.25000000
## 20 0 55 7 0 0 0 0 1 0 9 12 0.34523810
## 21 0 13 21 0 0 0 0 0 0 1 0 0.40000000
## 23 0 0 0 40 0 2 1 0 0 9 1 0.24528302
## 3 0 0 0 0 30 0 0 5 0 0 0 0.14285714
## 30 0 0 0 1 0 12 1 0 0 0 1 0.20000000
## 31 0 0 0 3 0 0 24 0 0 0 0 0.11111111
## 501 0 0 0 0 6 0 0 109 0 0 2 0.06837607
## 502 0 4 0 0 0 0 0 0 9 0 1 0.35714286
## 503 0 8 1 6 0 0 0 0 0 137 6 0.13291139
## 7 0 7 0 0 0 1 0 3 0 5 80 0.16666667map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN ProfileC ns5 45
mn="All_pRAN_ProfileC_NS5_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_ProfileC_NS5_pen45.shp", layer: "All_pRAN_ProfileC_NS5_pen45"
## with 936 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 936 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 19.12%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 3 0 0 0 2 0 0 0 0 0 0 0.40000000
## 20 0 83 9 0 0 0 0 1 0 13 15 0.31404959
## 21 0 11 33 0 0 0 0 0 1 3 0 0.31250000
## 23 0 0 0 52 0 7 3 0 0 14 2 0.33333333
## 3 1 0 0 0 42 0 0 9 0 0 0 0.19230769
## 30 0 0 0 5 0 14 0 0 0 0 3 0.36363636
## 31 0 0 0 4 0 0 37 0 0 0 0 0.09756098
## 501 0 0 0 0 9 0 0 164 1 0 1 0.06285714
## 502 0 4 2 0 0 0 0 2 11 0 1 0.45000000
## 503 0 11 2 7 0 0 0 0 0 210 5 0.10638298
## 7 0 11 0 3 0 3 0 6 0 8 108 0.22302158map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")ns7
SNA ProfileC ns7 15
mn="All_pSNA_ProfileC_NS7_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_ProfileC_NS7_pen15.shp", layer: "All_pSNA_ProfileC_NS7_pen15"
## with 122 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 122 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 31.15%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.0000000
## 20 0 10 1 0 0 0 0 0 0 3 3 0.4117647
## 21 0 3 1 0 0 0 0 0 0 0 0 0.7500000
## 23 0 0 0 9 0 0 0 0 0 1 0 0.1000000
## 3 0 0 0 0 4 0 0 2 0 0 0 0.3333333
## 30 0 0 0 1 0 2 0 0 0 0 0 0.3333333
## 31 0 0 0 0 0 0 4 0 0 0 0 0.0000000
## 501 0 1 0 0 1 0 0 21 0 0 1 0.1250000
## 502 0 1 0 0 0 0 0 2 0 0 1 1.0000000
## 503 0 5 1 1 0 0 0 1 0 19 2 0.3448276
## 7 0 2 0 1 0 0 0 1 0 2 14 0.3000000map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA ProfileC ns7 30
mn="All_pSNA_ProfileC_NS7_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_ProfileC_NS7_pen30.shp", layer: "All_pSNA_ProfileC_NS7_pen30"
## with 215 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 215 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 21.4%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.00000000
## 20 0 17 2 0 0 0 0 0 0 3 3 0.32000000
## 21 0 4 3 0 0 0 0 0 0 0 0 0.57142857
## 23 0 0 0 13 0 0 0 0 0 5 0 0.27777778
## 3 0 0 0 0 9 0 0 3 0 0 0 0.25000000
## 30 0 0 0 2 0 2 0 0 0 0 0 0.50000000
## 31 0 0 0 0 0 0 7 0 0 0 0 0.00000000
## 501 0 0 0 0 3 0 0 43 0 0 1 0.08510638
## 502 0 1 0 0 1 0 0 1 2 1 0 0.66666667
## 503 0 2 1 1 0 0 0 0 0 44 5 0.16981132
## 7 0 0 0 1 0 0 0 2 0 3 29 0.17142857map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA ProfileC ns7 45
mn="All_pSNA_ProfileC_NS7_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_ProfileC_NS7_pen45.shp", layer: "All_pSNA_ProfileC_NS7_pen45"
## with 307 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 307 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 15.96%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 2 0 0 0 0 0 0 0 0 0 0 0.00000000
## 20 0 21 1 0 0 0 0 0 1 5 3 0.32258065
## 21 0 4 4 0 0 0 0 0 0 0 0 0.50000000
## 23 0 0 0 21 0 0 0 0 0 5 0 0.19230769
## 3 0 0 0 0 15 0 0 3 0 0 0 0.16666667
## 30 0 0 0 1 0 3 1 0 0 0 0 0.40000000
## 31 0 0 0 0 0 0 11 0 0 0 0 0.00000000
## 501 0 0 0 0 2 0 0 67 0 0 1 0.04285714
## 502 0 1 0 0 0 0 0 3 1 1 2 0.87500000
## 503 0 2 1 3 0 0 0 0 0 70 3 0.11392405
## 7 0 1 0 1 0 0 0 1 0 3 43 0.12244898map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN ProfileC ns7 15
mn="All_pRAN_ProfileC_NS7_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_ProfileC_NS7_pen15.shp", layer: "All_pRAN_ProfileC_NS7_pen15"
## with 122 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 122 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 27.87%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.00000000
## 20 0 11 0 0 0 0 0 0 0 1 5 0.35294118
## 21 0 2 0 0 0 0 0 0 2 0 0 1.00000000
## 23 0 0 0 5 0 2 0 0 0 3 0 0.50000000
## 3 0 0 0 0 5 0 0 1 0 0 0 0.16666667
## 30 0 0 0 0 0 2 0 0 0 1 0 0.33333333
## 31 0 0 0 1 0 0 3 0 0 0 0 0.25000000
## 501 0 0 0 0 0 0 0 22 1 0 1 0.08333333
## 502 0 1 0 0 0 0 0 1 2 0 0 0.50000000
## 503 0 1 1 3 0 0 0 0 0 23 1 0.20689655
## 7 0 3 0 0 0 0 0 1 0 1 15 0.25000000map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN ProfileC ns7 30
mn="All_pRAN_ProfileC_NS7_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_ProfileC_NS7_pen30.shp", layer: "All_pRAN_ProfileC_NS7_pen30"
## with 215 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 215 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 18.14%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.00000000
## 20 0 19 1 0 0 0 0 0 0 2 3 0.24000000
## 21 0 2 5 0 0 0 0 0 0 0 0 0.28571429
## 23 0 0 0 12 0 2 0 0 0 4 0 0.33333333
## 3 0 0 0 0 8 0 0 4 0 0 0 0.33333333
## 30 0 0 0 2 0 2 0 0 0 0 0 0.50000000
## 31 0 0 0 0 0 0 7 0 0 0 0 0.00000000
## 501 0 1 0 0 2 0 0 44 0 0 0 0.06382979
## 502 0 2 0 0 1 0 0 2 1 0 0 0.83333333
## 503 0 2 0 2 0 0 0 0 0 48 1 0.09433962
## 7 0 2 0 1 0 0 0 0 0 2 30 0.14285714map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN ProfileC ns7 45
mn="All_pRAN_ProfileC_NS7_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_ProfileC_NS7_pen45.shp", layer: "All_pRAN_ProfileC_NS7_pen45"
## with 307 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 307 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 15.64%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 1 0 0 0 1 0 0 0 0 0 0 0.50000000
## 20 0 21 1 0 0 0 0 0 0 2 7 0.32258065
## 21 0 3 5 0 0 0 0 0 0 0 0 0.37500000
## 23 0 0 0 19 0 2 1 0 0 4 0 0.26923077
## 3 0 0 0 0 15 0 0 3 0 0 0 0.16666667
## 30 0 0 0 3 0 2 0 0 0 0 0 0.60000000
## 31 0 0 0 0 0 0 11 0 0 0 0 0.00000000
## 501 0 1 0 0 2 0 0 67 0 0 0 0.04285714
## 502 0 1 0 0 0 0 0 4 3 0 0 0.62500000
## 503 0 1 1 2 0 0 0 0 0 74 1 0.06329114
## 7 0 6 0 0 0 0 0 0 0 2 41 0.16326531map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")TWI -ok
ns3
SNA TWI ns3 15
mn="All_pSNA_TWI_NS3_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_TWI_NS3_pen15.shp", layer: "All_pSNA_TWI_NS3_pen15"
## with 985 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 983 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 22.69%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 6 0 0 0 0 0 0 0 0 0 0 0.00000000
## 20 0 87 15 1 0 0 0 1 0 31 16 0.42384106
## 21 0 23 23 0 0 0 0 0 0 7 2 0.58181818
## 23 0 0 0 90 0 2 1 0 0 7 1 0.10891089
## 3 0 0 0 0 36 0 0 14 0 0 0 0.28000000
## 30 0 0 0 5 0 23 3 0 0 0 2 0.30303030
## 31 0 0 0 0 0 1 41 0 0 0 0 0.02380952
## 501 0 1 0 0 10 0 0 141 2 0 3 0.10191083
## 502 0 1 0 0 0 0 0 3 14 0 2 0.30000000
## 503 0 25 4 9 0 0 0 0 0 187 3 0.17982456
## 7 0 13 1 1 0 1 0 6 1 5 112 0.20000000map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA TWI ns3 30
mn="All_pSNA_TWI_NS3_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_TWI_NS3_pen30.shp", layer: "All_pSNA_TWI_NS3_pen30"
## with 1921 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 1919 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 21.57%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 9 0 0 0 2 0 0 0 0 0 0 0.1818182
## 20 0 183 24 2 0 0 0 1 0 53 31 0.3775510
## 21 0 38 52 0 0 0 0 0 1 12 1 0.5000000
## 23 0 1 0 166 0 7 5 0 0 14 0 0.1398964
## 3 3 0 0 0 68 0 0 28 0 0 0 0.3131313
## 30 0 0 0 13 0 40 3 0 0 0 7 0.3650794
## 31 0 0 0 7 0 1 74 0 0 1 0 0.1084337
## 501 0 1 0 0 26 0 0 276 2 0 8 0.1182109
## 502 0 3 0 0 0 0 0 2 28 0 5 0.2631579
## 503 0 36 6 19 0 0 0 0 0 385 3 0.1425390
## 7 0 18 1 1 0 4 0 12 4 8 224 0.1764706map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA TWI ns3 45
mn="All_pSNA_TWI_NS3_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_TWI_NS3_pen45.shp", layer: "All_pSNA_TWI_NS3_pen45"
## with 2859 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 2857 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 20.48%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 13 0 0 0 3 0 0 0 0 0 0 0.1875000
## 20 0 275 33 2 0 0 0 4 3 72 47 0.3692661
## 21 0 50 90 0 0 0 0 0 3 12 1 0.4230769
## 23 0 1 0 239 0 11 9 0 0 22 4 0.1643357
## 3 3 0 0 0 102 0 0 42 0 0 0 0.3061224
## 30 0 0 0 12 0 69 4 0 0 0 6 0.2417582
## 31 0 0 0 10 0 4 110 0 0 0 0 0.1129032
## 501 0 2 0 0 31 0 0 414 1 0 21 0.1172708
## 502 0 2 1 0 0 0 0 2 41 0 8 0.2407407
## 503 0 52 6 24 0 1 0 0 0 579 10 0.1383929
## 7 0 22 1 1 0 4 0 21 4 13 340 0.1625616map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN TWI ns3 15
mn="All_pRAN_TWI_NS3_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_TWI_NS3_pen15.shp", layer: "All_pRAN_TWI_NS3_pen15"
## with 985 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 983 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 27.47%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 5 0 0 0 1 0 0 0 0 0 0 0.1666667
## 20 0 86 21 1 0 0 0 3 0 26 14 0.4304636
## 21 0 25 27 0 0 0 0 1 1 1 0 0.5090909
## 23 0 0 0 65 0 9 3 0 0 19 5 0.3564356
## 3 0 0 0 0 37 0 0 13 0 0 0 0.2600000
## 30 0 0 0 11 0 17 4 0 0 0 1 0.4848485
## 31 0 0 0 8 0 2 32 0 0 0 0 0.2380952
## 501 0 0 0 0 11 0 0 138 0 0 8 0.1210191
## 502 0 5 0 0 0 0 0 4 9 1 1 0.5500000
## 503 0 16 0 11 0 1 0 0 1 193 6 0.1535088
## 7 0 12 0 2 0 3 0 8 2 9 104 0.2571429map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN TWI ns3 30
mn="All_pRAN_TWI_NS3_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_TWI_NS3_pen30.shp", layer: "All_pRAN_TWI_NS3_pen30"
## with 1921 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 1918 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 21.64%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 8 0 0 0 2 0 0 0 0 0 0 0.2000000
## 20 0 203 23 0 0 0 0 2 0 37 29 0.3095238
## 21 0 45 52 0 0 0 0 1 1 4 1 0.5000000
## 23 0 0 0 139 0 12 6 0 0 27 9 0.2797927
## 3 2 0 0 0 78 0 0 19 0 0 0 0.2121212
## 30 0 0 0 14 0 45 3 0 0 0 1 0.2857143
## 31 0 0 0 11 0 3 69 0 0 0 0 0.1686747
## 501 0 5 0 0 15 0 0 288 1 0 4 0.0798722
## 502 0 4 0 0 0 0 0 1 30 0 3 0.2105263
## 503 0 35 3 16 0 0 0 0 0 385 10 0.1425390
## 7 0 23 1 4 0 3 0 16 2 17 206 0.2426471map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN TWI ns3 45
mn="All_pRAN_TWI_NS3_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_TWI_NS3_pen45.shp", layer: "All_pRAN_TWI_NS3_pen45"
## with 2859 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 2855 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 19.02%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 13 0 0 0 1 0 0 0 0 0 0 0.07142857
## 20 0 305 36 2 0 0 0 2 2 61 28 0.30045872
## 21 0 53 94 0 0 0 0 0 2 6 1 0.39743590
## 23 0 0 0 212 0 17 9 0 0 36 12 0.25874126
## 3 1 0 0 0 117 0 0 29 0 0 0 0.20408163
## 30 0 0 0 20 0 65 3 0 0 0 3 0.28571429
## 31 0 0 0 14 0 3 107 0 0 0 0 0.13709677
## 501 0 3 0 0 14 0 0 440 0 0 12 0.06183369
## 502 0 2 0 0 0 0 0 0 50 2 0 0.07407407
## 503 0 56 4 27 0 0 0 0 1 572 12 0.14880952
## 7 0 28 0 6 0 3 0 10 3 19 337 0.16995074map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")ns5
SNA TWI ns5 15
mn="All_pSNA_TWI_NS5_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_TWI_NS5_pen15.shp", layer: "All_pSNA_TWI_NS5_pen15"
## with 337 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 337 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 21.36%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 2 0 0 0 0 0 0 0 0 0 0 0.00000000
## 20 0 24 4 0 0 0 0 0 0 10 6 0.45454545
## 21 0 6 13 1 0 0 0 0 0 2 0 0.40909091
## 23 0 0 0 24 0 1 1 0 0 3 0 0.17241379
## 3 0 0 0 0 9 0 0 8 0 0 0 0.47058824
## 30 0 0 0 3 0 3 2 0 0 0 1 0.66666667
## 31 0 0 0 0 0 1 13 0 0 0 0 0.07142857
## 501 0 0 0 0 4 0 0 54 0 0 1 0.08474576
## 502 0 1 0 0 0 0 0 2 4 0 0 0.42857143
## 503 0 2 2 2 0 0 0 0 0 74 1 0.08641975
## 7 0 4 0 1 0 0 0 1 0 2 45 0.15094340map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA TWI ns5 30
mn="All_pSNA_TWI_NS5_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_TWI_NS5_pen30.shp", layer: "All_pSNA_TWI_NS5_pen30"
## with 634 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 634 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 22.71%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 3 0 0 0 1 0 0 0 0 0 0 0.2500000
## 20 0 40 11 0 0 0 0 0 0 18 13 0.5121951
## 21 0 10 17 0 0 0 0 0 2 3 0 0.4687500
## 23 0 0 0 44 0 3 1 0 0 6 0 0.1851852
## 3 1 0 0 0 19 0 0 14 0 0 0 0.4411765
## 30 0 0 0 3 0 7 3 0 0 0 2 0.5333333
## 31 0 0 0 2 0 3 22 0 0 1 0 0.2142857
## 501 0 0 0 0 10 0 0 105 0 0 2 0.1025641
## 502 0 1 0 0 0 0 0 1 10 0 2 0.2857143
## 503 0 6 2 7 0 0 0 0 0 137 3 0.1161290
## 7 0 6 1 0 0 0 0 3 1 2 86 0.1313131map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA TWI ns5 45
mn="All_pSNA_TWI_NS5_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_TWI_NS5_pen45.shp", layer: "All_pSNA_TWI_NS5_pen45"
## with 934 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 934 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 19.49%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 5 0 0 0 1 0 0 0 0 0 0 0.1666667
## 20 0 68 12 0 0 0 0 1 1 19 17 0.4237288
## 21 0 15 27 0 0 0 0 0 0 3 0 0.4000000
## 23 0 0 0 66 0 2 4 0 0 9 0 0.1851852
## 3 1 0 0 0 35 0 0 14 0 0 0 0.3000000
## 30 0 0 0 2 0 13 5 0 0 0 2 0.4090909
## 31 0 0 0 3 0 3 35 0 0 1 0 0.1666667
## 501 0 0 0 0 16 0 0 155 0 0 4 0.1142857
## 502 0 1 0 0 0 0 0 2 13 0 3 0.3157895
## 503 0 9 2 9 0 0 0 0 0 209 2 0.0952381
## 7 0 9 1 0 0 0 0 5 0 4 126 0.1310345map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN TWI ns5 15
mn="All_pRAN_TWI_NS5_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_TWI_NS5_pen15.shp", layer: "All_pRAN_TWI_NS5_pen15"
## with 337 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 337 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 23.15%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 1 0 0 0 1 0 0 0 0 0 0 0.5000000
## 20 0 26 6 0 0 0 0 1 1 2 8 0.4090909
## 21 0 8 13 0 0 0 0 0 0 1 0 0.4090909
## 23 0 0 0 22 0 3 1 0 0 3 0 0.2413793
## 3 0 0 0 0 11 0 0 5 1 0 0 0.3529412
## 30 0 0 0 4 0 4 0 0 0 0 1 0.5555556
## 31 0 0 0 0 0 0 14 0 0 0 0 0.0000000
## 501 0 0 0 0 6 0 0 52 0 0 1 0.1186441
## 502 0 2 1 0 0 0 0 1 3 0 0 0.5714286
## 503 0 2 1 6 0 0 0 0 0 69 3 0.1481481
## 7 0 6 0 0 0 0 0 1 0 2 44 0.1698113map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN TWI ns5 30
mn="All_pRAN_TWI_NS5_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_TWI_NS5_pen30.shp", layer: "All_pRAN_TWI_NS5_pen30"
## with 634 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 634 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 18.45%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 2 0 0 0 2 0 0 0 0 0 0 0.50000000
## 20 0 57 5 0 0 0 0 2 0 10 8 0.30487805
## 21 0 15 15 0 0 0 0 0 0 2 0 0.53125000
## 23 0 0 0 39 0 4 2 0 0 7 2 0.27777778
## 3 1 0 0 0 28 0 0 5 0 0 0 0.17647059
## 30 0 0 0 4 0 9 1 0 0 0 1 0.40000000
## 31 0 0 0 2 0 0 25 0 0 1 0 0.10714286
## 501 0 2 0 0 8 0 0 105 1 0 1 0.10256410
## 502 0 2 0 0 0 0 0 1 9 1 1 0.35714286
## 503 0 6 1 4 0 0 0 0 0 142 2 0.08387097
## 7 0 6 0 0 0 0 0 1 0 6 86 0.13131313map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN TWI ns5 45
mn="All_pRAN_TWI_NS5_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_TWI_NS5_pen45.shp", layer: "All_pRAN_TWI_NS5_pen45"
## with 934 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 934 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 16.81%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 5 0 0 0 1 0 0 0 0 0 0 0.16666667
## 20 0 79 8 0 0 0 0 1 1 15 14 0.33050847
## 21 0 16 28 0 0 0 0 0 0 1 0 0.37777778
## 23 0 0 0 61 0 6 2 0 0 9 3 0.24691358
## 3 0 0 0 0 41 0 0 9 0 0 0 0.18000000
## 30 0 0 0 7 0 12 1 0 0 0 2 0.45454545
## 31 0 0 0 0 0 1 41 0 0 0 0 0.02380952
## 501 0 1 0 0 5 0 0 165 0 0 4 0.05714286
## 502 0 2 0 0 0 0 0 3 13 0 1 0.31578947
## 503 0 6 0 8 0 0 0 0 0 213 4 0.07792208
## 7 0 8 0 1 0 1 0 6 1 9 119 0.17931034map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")ns7
SNA TWI ns7 15
mn="All_pSNA_TWI_NS7_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_TWI_NS7_pen15.shp", layer: "All_pSNA_TWI_NS7_pen15"
## with 122 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 122 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 27.87%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.00000000
## 20 0 8 1 0 0 0 0 0 0 5 3 0.52941176
## 21 0 2 2 0 0 0 0 0 0 0 0 0.50000000
## 23 0 0 0 6 0 2 0 0 0 3 0 0.45454545
## 3 0 0 0 0 5 0 0 2 0 0 0 0.28571429
## 30 0 0 0 2 0 1 0 0 0 0 0 0.66666667
## 31 0 0 0 0 0 0 4 0 0 0 0 0.00000000
## 501 0 0 0 0 1 0 0 22 0 0 0 0.04347826
## 502 0 0 0 0 0 0 0 3 0 0 0 1.00000000
## 503 0 3 0 1 0 0 0 0 0 25 0 0.13793103
## 7 0 2 0 1 0 0 0 1 0 1 15 0.25000000map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA TWI ns7 30
mn="All_pSNA_TWI_NS7_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_TWI_NS7_pen30.shp", layer: "All_pSNA_TWI_NS7_pen30"
## with 216 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 216 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 18.52%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.00000000
## 20 0 12 2 0 0 0 0 0 0 7 5 0.53846154
## 21 0 2 4 0 0 0 0 0 0 0 0 0.33333333
## 23 0 0 0 16 0 1 0 0 0 3 0 0.20000000
## 3 0 0 0 0 10 0 0 3 0 0 0 0.23076923
## 30 0 0 0 1 0 3 0 0 0 0 0 0.25000000
## 31 0 0 0 1 0 0 6 0 0 0 0 0.14285714
## 501 0 0 0 0 2 0 0 43 0 0 0 0.04444444
## 502 0 0 0 0 0 0 0 2 3 0 0 0.40000000
## 503 0 4 0 2 0 0 0 0 0 48 0 0.11111111
## 7 0 2 0 0 0 0 0 0 0 2 31 0.11428571map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA TWI ns7 45
mn="All_pSNA_TWI_NS7_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_TWI_NS7_pen45.shp", layer: "All_pSNA_TWI_NS7_pen45"
## with 310 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 310 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 19.35%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 2 0 0 0 0 0 0 1.00000000
## 20 0 19 1 0 0 0 0 0 0 7 7 0.44117647
## 21 0 3 5 0 0 0 0 0 0 0 0 0.37500000
## 23 0 0 0 20 0 1 1 0 0 5 0 0.25925926
## 3 1 0 0 0 11 0 0 7 0 0 0 0.42105263
## 30 0 0 0 1 0 3 1 0 0 0 0 0.40000000
## 31 0 0 0 0 0 1 10 0 0 0 0 0.09090909
## 501 0 0 0 0 3 0 0 62 0 0 2 0.07462687
## 502 0 0 0 0 0 0 0 3 4 0 0 0.42857143
## 503 0 3 0 1 0 0 0 0 0 73 2 0.07594937
## 7 0 3 0 0 0 0 0 2 0 3 43 0.15686275map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN TWI ns7 15
mn="All_pRAN_TWI_NS7_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_TWI_NS7_pen15.shp", layer: "All_pRAN_TWI_NS7_pen15"
## with 122 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 122 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 26.23%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.0000000
## 20 0 11 0 0 0 0 0 0 0 3 3 0.3529412
## 21 0 2 2 0 0 0 0 0 0 0 0 0.5000000
## 23 0 0 0 9 0 1 1 0 0 0 0 0.1818182
## 3 0 0 0 0 5 0 0 2 0 0 0 0.2857143
## 30 0 0 0 2 0 0 0 0 0 0 1 1.0000000
## 31 0 0 0 1 0 0 3 0 0 0 0 0.2500000
## 501 0 0 0 0 2 0 0 20 0 0 1 0.1304348
## 502 0 1 0 0 0 0 0 1 1 0 0 0.6666667
## 503 0 3 0 0 0 0 0 0 0 26 0 0.1034483
## 7 0 3 0 0 0 1 0 1 0 2 13 0.3500000map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN TWI ns7 30
mn="All_pRAN_TWI_NS7_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_TWI_NS7_pen30.shp", layer: "All_pRAN_TWI_NS7_pen30"
## with 216 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 216 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 19.44%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.00000000
## 20 0 16 2 0 0 0 0 0 0 3 5 0.38461538
## 21 0 3 2 0 0 0 0 0 1 0 0 0.66666667
## 23 0 0 0 19 0 1 0 0 0 0 0 0.05000000
## 3 0 0 0 0 10 0 0 3 0 0 0 0.23076923
## 30 0 0 0 2 0 1 1 0 0 0 0 0.75000000
## 31 0 0 0 0 0 1 6 0 0 0 0 0.14285714
## 501 0 0 0 0 1 0 0 42 1 0 1 0.06666667
## 502 0 1 0 0 0 0 0 3 1 0 0 0.80000000
## 503 0 2 0 2 0 0 0 0 0 48 2 0.11111111
## 7 0 3 0 1 0 0 0 1 0 1 29 0.17142857map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN TWI ns7 45
mn="All_pRAN_TWI_NS7_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_TWI_NS7_pen45.shp", layer: "All_pRAN_TWI_NS7_pen45"
## with 310 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 310 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 14.52%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 2 0 0 0 0 0 0 0 0 0 0 0.00000000
## 20 0 22 2 0 0 0 0 0 1 5 4 0.35294118
## 21 0 3 5 0 0 0 0 0 0 0 0 0.37500000
## 23 0 0 0 21 0 2 1 0 0 3 0 0.22222222
## 3 0 0 0 0 17 0 0 2 0 0 0 0.10526316
## 30 0 0 0 3 0 2 0 0 0 0 0 0.60000000
## 31 0 0 0 0 0 0 11 0 0 0 0 0.00000000
## 501 0 0 0 0 2 0 0 64 0 0 1 0.04477612
## 502 0 1 0 0 0 0 0 2 4 0 0 0.42857143
## 503 0 3 1 2 0 0 0 0 0 71 2 0.10126582
## 7 0 1 0 1 0 0 0 0 0 3 46 0.09803922map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")LSF -ok
ns3
SNA LSF ns3 15
mn="All_pSNA_LSF_NS3_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_LSF_NS3_pen15.shp", layer: "All_pSNA_LSF_NS3_pen15"
## with 987 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 985 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 19.49%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 2 0 0 0 4 0 0 0 0 0 0 0.66666667
## 20 0 114 7 0 0 0 0 1 2 14 17 0.26451613
## 21 0 20 36 0 0 0 0 0 1 1 0 0.37931034
## 23 0 0 0 75 0 3 5 0 0 17 0 0.25000000
## 3 0 0 0 0 51 0 0 1 0 0 0 0.01923077
## 30 0 0 0 8 0 20 2 0 0 0 2 0.37500000
## 31 0 0 0 3 0 0 38 0 0 0 0 0.07317073
## 501 0 1 0 0 9 0 0 142 1 0 5 0.10126582
## 502 0 2 0 0 0 0 0 1 14 2 2 0.33333333
## 503 0 12 0 13 0 0 0 0 2 196 4 0.13656388
## 7 0 16 0 1 0 1 0 6 0 6 105 0.22222222map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA LSF ns3 30
mn="All_pSNA_LSF_NS3_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_LSF_NS3_pen30.shp", layer: "All_pSNA_LSF_NS3_pen30"
## with 1926 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 1924 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 18.61%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 6 0 0 0 6 0 0 0 0 0 0 0.50000000
## 20 0 209 17 0 0 0 0 7 2 37 31 0.31023102
## 21 0 37 65 0 0 0 0 1 2 3 0 0.39814815
## 23 0 0 0 150 0 6 6 0 0 27 2 0.21465969
## 3 3 0 0 0 88 0 0 12 0 0 0 0.14563107
## 30 0 0 0 7 0 49 2 0 0 0 3 0.19672131
## 31 0 0 0 4 0 0 76 0 0 1 0 0.06172840
## 501 0 2 0 0 8 0 0 298 0 0 6 0.05095541
## 502 0 6 1 0 0 0 0 2 25 1 4 0.35897436
## 503 0 25 1 15 0 0 0 0 2 391 12 0.12331839
## 7 0 29 0 0 0 3 0 16 1 8 209 0.21428571map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA LSF ns3 45
mn="All_pSNA_LSF_NS3_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_LSF_NS3_pen45.shp", layer: "All_pSNA_LSF_NS3_pen45"
## with 2858 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 2856 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 16.84%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 10 0 0 0 7 0 0 0 0 0 0 0.41176471
## 20 0 331 30 0 0 0 0 4 4 43 35 0.25950783
## 21 0 52 98 0 0 0 0 2 2 4 0 0.37974684
## 23 0 0 0 231 0 9 5 0 0 35 2 0.18085106
## 3 2 0 0 0 133 0 0 18 0 0 0 0.13071895
## 30 0 0 0 14 0 66 3 0 0 0 6 0.25842697
## 31 0 0 0 7 0 4 109 0 0 1 0 0.09917355
## 501 0 2 0 0 13 0 0 444 0 0 12 0.05732484
## 502 0 4 0 0 0 0 0 2 45 2 4 0.21052632
## 503 0 40 2 25 0 1 0 0 2 585 11 0.12162162
## 7 0 32 1 0 0 3 0 20 1 15 323 0.18227848map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN LSF ns3 15
mn="All_pRAN_LSF_NS3_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_LSF_NS3_pen15.shp", layer: "All_pRAN_LSF_NS3_pen15"
## with 987 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 985 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 29.04%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 6 0 0 0 0 0 0 0 0 0 0 0.00000000
## 20 0 90 25 0 0 0 0 1 1 22 16 0.41935484
## 21 0 37 18 0 0 0 0 0 1 1 1 0.68965517
## 23 0 0 0 64 0 5 5 0 0 21 5 0.36000000
## 3 1 0 0 0 44 0 0 7 0 0 0 0.15384615
## 30 0 0 0 9 0 18 4 0 0 1 0 0.43750000
## 31 0 0 0 7 0 3 30 0 0 1 0 0.26829268
## 501 0 1 0 0 8 0 0 146 0 0 3 0.07594937
## 502 0 4 1 0 0 0 0 5 9 1 1 0.57142857
## 503 0 17 5 16 0 1 0 0 0 180 8 0.20704846
## 7 0 19 1 2 0 2 0 5 1 11 94 0.30370370map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN LSF ns3 30
mn="All_pRAN_LSF_NS3_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_LSF_NS3_pen30.shp", layer: "All_pRAN_LSF_NS3_pen30"
## with 1926 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 1922 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 23.26%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 10 0 0 0 0 0 0 0 0 0 0 0.00000000
## 20 0 207 23 1 0 1 0 2 1 45 23 0.31683168
## 21 0 43 61 0 0 0 0 0 0 4 0 0.43518519
## 23 0 0 0 137 0 11 5 0 0 26 12 0.28272251
## 3 1 0 0 0 79 0 0 23 0 0 0 0.23300971
## 30 0 0 0 17 0 37 3 0 0 0 4 0.39344262
## 31 0 0 0 11 0 5 65 0 0 0 0 0.19753086
## 501 0 1 0 0 19 0 0 286 0 0 8 0.08917197
## 502 0 5 1 0 0 0 0 2 27 1 3 0.30769231
## 503 0 35 2 21 0 0 0 0 1 376 11 0.15695067
## 7 0 26 1 6 0 1 0 19 2 21 190 0.28571429map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN LSF ns3 45
mn="All_pRAN_LSF_NS3_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_LSF_NS3_pen45.shp", layer: "All_pRAN_LSF_NS3_pen45"
## with 2858 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 2853 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 20.15%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 14 0 0 0 0 0 0 0 0 0 0 0.00000000
## 20 0 323 31 0 0 0 0 3 1 62 27 0.27740492
## 21 0 54 92 0 0 0 0 0 0 11 1 0.41772152
## 23 0 1 0 218 0 20 10 0 0 24 9 0.22695035
## 3 2 0 0 0 118 0 0 33 0 0 0 0.22875817
## 30 0 0 0 22 0 60 3 0 0 1 3 0.32584270
## 31 0 0 0 13 0 7 99 0 0 2 0 0.18181818
## 501 0 5 0 0 28 0 0 428 1 0 9 0.09129512
## 502 0 7 1 0 0 0 0 0 44 1 4 0.22807018
## 503 0 37 6 35 0 1 0 0 0 572 15 0.14114114
## 7 0 34 1 3 0 4 0 18 1 24 310 0.21518987map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")ns5
SNA LSF ns5 15
mn="All_pSNA_LSF_NS5_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_LSF_NS5_pen15.shp", layer: "All_pSNA_LSF_NS5_pen15"
## with 334 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 334 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 20.06%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 1 0 0 0 1 0 0 0 0 0 0 0.50000000
## 20 0 26 3 0 0 0 0 2 0 7 6 0.40909091
## 21 0 6 13 0 0 0 0 0 0 1 1 0.38095238
## 23 0 0 0 22 0 2 0 0 0 5 0 0.24137931
## 3 0 0 0 0 19 0 0 0 0 0 0 0.00000000
## 30 0 0 0 2 0 6 1 0 0 0 0 0.33333333
## 31 0 0 0 0 0 1 11 0 0 0 0 0.08333333
## 501 0 2 0 0 0 0 0 56 0 0 2 0.06666667
## 502 0 2 0 0 0 0 0 0 4 0 1 0.42857143
## 503 0 3 0 1 0 0 0 0 0 72 3 0.08860759
## 7 0 9 0 0 0 1 0 2 0 3 37 0.28846154map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA LSF ns5 30
mn="All_pSNA_LSF_NS5_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_LSF_NS5_pen30.shp", layer: "All_pSNA_LSF_NS5_pen30"
## with 629 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 629 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 18.12%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 1 0 0 0 3 0 0 0 0 0 0 0.75000000
## 20 0 47 7 0 0 0 0 4 0 9 14 0.41975309
## 21 0 11 18 0 0 0 0 1 1 1 0 0.43750000
## 23 0 0 0 41 0 4 1 0 0 7 1 0.24074074
## 3 0 0 0 0 36 0 0 1 0 0 0 0.02702703
## 30 0 0 0 3 0 11 1 0 0 0 0 0.26666667
## 31 0 0 0 0 0 2 22 0 0 0 0 0.08333333
## 501 0 2 0 0 3 0 0 114 0 0 1 0.05000000
## 502 0 4 0 0 0 0 0 0 10 0 0 0.28571429
## 503 0 9 0 4 0 0 0 0 0 137 2 0.09868421
## 7 0 11 0 0 0 1 0 1 0 5 78 0.18750000map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA LSF ns5 45
mn="All_pSNA_LSF_NS5_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_LSF_NS5_pen45.shp", layer: "All_pSNA_LSF_NS5_pen45"
## with 927 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 927 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 16.07%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 1 0 0 0 4 0 0 0 0 0 0 0.80000000
## 20 0 73 9 0 0 0 0 5 0 15 16 0.38135593
## 21 0 14 27 0 0 0 0 1 0 1 1 0.38636364
## 23 0 0 0 68 0 4 2 0 0 7 0 0.16049383
## 3 1 0 0 0 53 0 0 1 0 0 0 0.03636364
## 30 0 0 0 4 0 17 0 0 0 0 1 0.22727273
## 31 0 0 0 0 0 0 36 0 0 0 0 0.00000000
## 501 0 1 0 0 5 0 0 168 0 0 4 0.05617978
## 502 0 5 0 0 0 0 0 0 15 0 1 0.28571429
## 503 0 12 0 5 0 0 0 0 1 202 6 0.10619469
## 7 0 12 0 0 0 1 0 3 0 7 118 0.16312057map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN LSF ns5 15
mn="All_pRAN_LSF_NS5_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_LSF_NS5_pen15.shp", layer: "All_pRAN_LSF_NS5_pen15"
## with 334 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 334 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 25.45%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 1 0 0 0 1 0 0 0 0 0 0 0.5000000
## 20 0 25 5 0 0 0 0 0 0 6 8 0.4318182
## 21 0 8 12 0 0 0 0 0 0 1 0 0.4285714
## 23 0 0 0 19 0 3 1 0 0 4 2 0.3448276
## 3 1 0 0 0 15 0 0 3 0 0 0 0.2105263
## 30 0 0 0 6 0 2 1 0 0 0 0 0.7777778
## 31 0 0 0 2 0 1 9 0 0 0 0 0.2500000
## 501 0 0 0 0 4 0 0 54 1 0 1 0.1000000
## 502 0 1 1 0 0 0 0 2 2 0 1 0.7142857
## 503 0 5 1 3 0 0 0 0 0 70 0 0.1139241
## 7 0 6 0 2 0 0 0 1 0 3 40 0.2307692map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN LSF ns5 30
mn="All_pRAN_LSF_NS5_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_LSF_NS5_pen30.shp", layer: "All_pRAN_LSF_NS5_pen30"
## with 629 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 629 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 21.14%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 2 0 0 0 2 0 0 0 0 0 0 0.5000000
## 20 0 56 6 0 0 0 0 0 0 8 11 0.3086420
## 21 0 14 17 0 0 0 0 0 0 1 0 0.4687500
## 23 0 0 0 36 0 4 3 0 0 9 2 0.3333333
## 3 0 0 0 0 29 0 0 8 0 0 0 0.2162162
## 30 0 0 0 5 0 7 2 0 0 0 1 0.5333333
## 31 0 0 0 4 0 0 20 0 0 0 0 0.1666667
## 501 0 1 0 0 10 0 0 108 1 0 0 0.1000000
## 502 0 3 0 0 0 0 0 1 10 0 0 0.2857143
## 503 0 4 3 6 0 0 0 0 0 134 5 0.1184211
## 7 0 9 0 2 0 0 0 2 0 6 77 0.1979167map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN LSF ns5 45
mn="All_pRAN_LSF_NS5_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_LSF_NS5_pen45.shp", layer: "All_pRAN_LSF_NS5_pen45"
## with 927 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 927 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 17.91%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 4 0 0 0 1 0 0 0 0 0 0 0.20000000
## 20 0 77 8 0 0 0 0 1 0 14 18 0.34745763
## 21 0 16 27 0 0 0 0 0 0 1 0 0.38636364
## 23 0 0 0 58 0 7 2 0 0 12 2 0.28395062
## 3 0 0 0 0 43 0 0 12 0 0 0 0.21818182
## 30 0 0 0 5 0 15 0 0 0 0 2 0.31818182
## 31 0 0 0 4 0 1 30 0 0 0 1 0.16666667
## 501 0 2 0 0 6 0 0 168 0 0 2 0.05617978
## 502 0 1 0 0 0 0 0 3 15 0 2 0.28571429
## 503 0 8 1 7 0 0 0 0 0 204 6 0.09734513
## 7 0 11 0 1 0 1 0 0 0 8 120 0.14893617map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")ns7
SNA LSF ns7 15
mn="All_pSNA_LSF_NS7_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_LSF_NS7_pen15.shp", layer: "All_pSNA_LSF_NS7_pen15"
## with 121 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 121 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 28.93%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.00000000
## 20 0 11 2 0 0 0 0 0 0 2 3 0.38888889
## 21 0 2 2 0 0 0 0 0 0 0 0 0.50000000
## 23 0 0 0 7 0 1 0 0 0 3 0 0.36363636
## 3 0 0 0 0 6 0 0 1 0 0 0 0.14285714
## 30 0 0 0 2 0 0 0 0 0 0 0 1.00000000
## 31 0 0 0 0 0 0 4 0 0 0 0 0.00000000
## 501 0 0 0 0 0 0 0 22 0 0 1 0.04347826
## 502 0 1 0 0 1 0 0 0 0 0 1 1.00000000
## 503 0 1 0 3 0 0 0 0 0 24 0 0.14285714
## 7 0 7 0 0 0 0 0 1 0 2 10 0.50000000map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA LSF ns7 30
mn="All_pSNA_LSF_NS7_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_LSF_NS7_pen30.shp", layer: "All_pSNA_LSF_NS7_pen30"
## with 217 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 217 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 17.97%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 2 0 0 0 0 0 0 1.00000000
## 20 0 17 0 0 0 0 0 0 0 6 5 0.39285714
## 21 0 3 2 0 0 0 0 0 0 0 0 0.60000000
## 23 0 0 0 15 0 1 0 0 0 3 0 0.21052632
## 3 0 0 0 0 14 0 0 0 0 0 0 0.00000000
## 30 0 0 0 1 0 2 0 0 0 0 0 0.33333333
## 31 0 0 0 0 0 0 7 0 0 0 0 0.00000000
## 501 0 0 0 0 0 0 0 45 0 0 1 0.02173913
## 502 0 1 0 0 1 0 0 1 2 0 0 0.60000000
## 503 0 2 0 2 0 0 0 0 0 48 1 0.09433962
## 7 0 6 0 0 0 0 0 1 0 2 26 0.25714286map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA LSF ns7 45
mn="All_pSNA_LSF_NS7_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_LSF_NS7_pen45.shp", layer: "All_pSNA_LSF_NS7_pen45"
## with 309 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 309 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 18.77%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 2 0 0 0 0 0 0 1.00000000
## 20 0 21 2 0 0 0 0 0 1 6 6 0.41666667
## 21 0 2 6 0 0 0 0 0 0 0 0 0.25000000
## 23 0 0 0 19 0 0 0 0 0 8 0 0.29629630
## 3 0 0 0 0 20 0 0 0 0 0 0 0.00000000
## 30 0 0 0 1 0 3 0 0 0 0 0 0.25000000
## 31 0 0 0 1 0 0 10 0 0 0 0 0.09090909
## 501 0 0 0 0 0 0 0 65 1 0 2 0.04411765
## 502 0 0 0 0 0 0 0 2 4 0 1 0.42857143
## 503 0 3 0 5 0 0 0 0 0 66 3 0.14285714
## 7 0 7 0 0 0 0 0 1 0 4 37 0.24489796map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN LSF ns7 15
mn="All_pRAN_LSF_NS7_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_LSF_NS7_pen15.shp", layer: "All_pRAN_LSF_NS7_pen15"
## with 121 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 121 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 30.58%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.00000000
## 20 0 12 1 0 0 0 0 0 0 2 3 0.33333333
## 21 0 3 0 0 0 0 0 0 1 0 0 1.00000000
## 23 0 0 0 5 0 2 1 0 0 3 0 0.54545455
## 3 0 0 0 0 5 0 0 2 0 0 0 0.28571429
## 30 0 0 0 1 0 0 0 0 0 1 0 1.00000000
## 31 0 0 0 3 0 0 1 0 0 0 0 0.75000000
## 501 0 0 0 0 1 0 0 21 1 0 0 0.08695652
## 502 0 1 1 0 1 0 0 0 0 0 0 1.00000000
## 503 0 2 0 2 0 0 0 0 0 23 1 0.17857143
## 7 0 1 0 0 0 0 0 0 0 2 17 0.15000000map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN LSF ns7 30
mn="All_pRAN_LSF_NS7_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_LSF_NS7_pen30.shp", layer: "All_pRAN_LSF_NS7_pen30"
## with 217 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 217 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 18.89%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 1 0 0 0 1 0 0 0 0 0 0 0.50000000
## 20 0 19 0 0 0 0 0 0 1 4 4 0.32142857
## 21 0 3 1 0 0 0 0 0 1 0 0 0.80000000
## 23 0 0 0 13 0 2 1 0 0 3 0 0.31578947
## 3 1 0 0 0 12 0 0 1 0 0 0 0.14285714
## 30 0 0 0 2 0 0 0 0 0 1 0 1.00000000
## 31 0 0 0 1 0 0 6 0 0 0 0 0.14285714
## 501 0 0 0 0 0 0 0 44 1 0 1 0.04347826
## 502 0 2 0 0 1 0 0 0 1 0 1 0.80000000
## 503 0 3 0 3 0 0 0 0 0 47 0 0.11320755
## 7 0 1 0 0 0 0 0 1 0 1 32 0.08571429map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN LSF ns7 45
mn="All_pRAN_LSF_NS7_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_LSF_NS7_pen45.shp", layer: "All_pRAN_LSF_NS7_pen45"
## with 309 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 309 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 15.53%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 1 0 0 0 1 0 0 0 0 0 0 0.50000000
## 20 0 28 1 0 0 0 0 0 0 4 3 0.22222222
## 21 0 2 4 0 0 0 0 0 2 0 0 0.50000000
## 23 0 0 0 18 0 3 0 0 0 6 0 0.33333333
## 3 0 0 0 0 18 0 0 2 0 0 0 0.10000000
## 30 0 0 0 4 0 0 0 0 0 0 0 1.00000000
## 31 0 0 0 2 0 0 9 0 0 0 0 0.18181818
## 501 0 0 0 0 3 0 0 64 0 0 1 0.05882353
## 502 0 0 0 0 0 0 0 2 4 0 1 0.42857143
## 503 0 4 0 1 0 0 0 0 0 70 2 0.09090909
## 7 0 2 0 0 0 0 0 1 0 1 45 0.08163265map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")VD -ok
ns3
SNA VD ns3 15
mn="All_pSNA_VD_NS3_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_VD_NS3_pen15.shp", layer: "All_pSNA_VD_NS3_pen15"
## with 1042 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 1040 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 16.83%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 3 0 0 0 3 0 0 0 0 0 0 0.50000000
## 20 0 125 7 0 0 0 0 1 0 19 10 0.22839506
## 21 0 22 34 0 0 0 0 1 0 3 0 0.43333333
## 23 0 0 0 78 0 7 0 0 0 11 6 0.23529412
## 3 1 0 0 0 45 0 0 9 0 0 0 0.18181818
## 30 0 0 0 8 0 25 0 0 0 0 0 0.24242424
## 31 0 0 0 6 0 1 36 0 0 0 0 0.16279070
## 501 0 0 0 0 8 0 0 161 0 0 1 0.05294118
## 502 0 4 0 0 0 0 0 3 13 0 2 0.40909091
## 503 0 13 0 6 0 0 0 0 0 215 9 0.11522634
## 7 0 8 0 0 0 0 0 2 0 4 130 0.09722222map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA VD ns3 30
mn="All_pSNA_VD_NS3_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_VD_NS3_pen30.shp", layer: "All_pSNA_VD_NS3_pen30"
## with 2045 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 2043 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 15.96%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 9 0 0 0 3 0 0 0 0 0 0 0.2500000
## 20 0 242 15 0 0 0 0 0 1 38 22 0.2389937
## 21 0 33 73 0 0 0 0 0 1 4 1 0.3482143
## 23 0 0 0 155 0 10 2 0 0 18 10 0.2051282
## 3 3 0 0 0 94 0 0 13 0 0 0 0.1454545
## 30 0 0 0 12 0 47 1 0 0 1 1 0.2419355
## 31 0 0 0 10 0 2 73 0 0 0 0 0.1411765
## 501 0 0 0 0 13 0 0 323 1 1 2 0.0500000
## 502 0 5 2 0 0 0 0 6 27 0 2 0.3571429
## 503 0 26 0 11 0 1 0 1 0 432 13 0.1074380
## 7 0 19 0 2 0 0 0 5 1 14 242 0.1448763map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA VD ns3 45
mn="All_pSNA_VD_NS3_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_VD_NS3_pen45.shp", layer: "All_pSNA_VD_NS3_pen45"
## with 3041 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 3035 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 15.62%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 12 0 0 0 3 0 0 0 0 0 0 0.20000000
## 20 0 357 24 0 0 0 0 3 0 55 31 0.24042553
## 21 0 56 104 0 0 0 0 0 0 5 1 0.37349398
## 23 0 0 0 229 0 18 8 0 0 21 13 0.20761246
## 3 3 0 0 0 140 0 0 22 0 0 0 0.15151515
## 30 0 0 0 18 0 66 2 0 0 1 3 0.26666667
## 31 0 0 0 12 0 3 112 0 0 0 0 0.11811024
## 501 0 1 0 0 17 0 0 489 1 0 2 0.04117647
## 502 0 4 2 0 0 0 0 7 45 0 3 0.26229508
## 503 0 42 0 17 0 1 0 2 0 646 15 0.10650069
## 7 0 27 0 3 0 2 0 7 1 18 361 0.13842482map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN VD ns3 15
mn="All_pRAN_VD_NS3_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_VD_NS3_pen15.shp", layer: "All_pRAN_VD_NS3_pen15"
## with 1042 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 1039 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 25.79%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 3 0 0 0 2 0 0 0 0 0 0 0.40000000
## 20 0 107 12 0 0 0 0 0 3 25 15 0.33950617
## 21 0 31 27 0 0 0 0 1 0 1 0 0.55000000
## 23 0 0 0 70 0 7 3 0 0 18 4 0.31372549
## 3 3 0 0 0 39 0 0 13 0 0 0 0.29090909
## 30 0 0 0 9 0 14 5 0 0 2 3 0.57575758
## 31 0 0 0 8 0 5 30 0 0 0 0 0.30232558
## 501 0 0 0 0 8 0 0 158 0 0 4 0.07058824
## 502 0 2 1 0 0 0 0 6 10 0 3 0.54545455
## 503 0 28 1 9 0 2 0 0 0 198 5 0.18518519
## 7 0 17 0 1 0 3 0 3 1 4 115 0.20138889map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN VD ns3 30
mn="All_pRAN_VD_NS3_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_VD_NS3_pen30.shp", layer: "All_pRAN_VD_NS3_pen30"
## with 2045 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 2042 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 21.25%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 10 0 0 0 1 0 0 0 0 0 0 0.09090909
## 20 0 211 31 0 0 0 0 6 2 50 18 0.33647799
## 21 0 44 63 0 0 0 0 1 1 3 0 0.43750000
## 23 0 0 0 146 0 11 4 0 0 24 10 0.25128205
## 3 0 0 0 0 88 0 0 22 0 0 0 0.20000000
## 30 0 0 0 16 0 38 5 0 0 1 2 0.38709677
## 31 0 0 0 12 0 4 69 0 0 0 0 0.18823529
## 501 0 1 1 0 19 0 0 310 0 0 9 0.08823529
## 502 0 6 0 0 0 0 0 3 31 0 2 0.26190476
## 503 0 37 3 13 0 0 0 0 0 417 14 0.13842975
## 7 0 25 0 4 0 2 0 11 3 13 225 0.20494700map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN VD ns3 45
mn="All_pRAN_VD_NS3_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_VD_NS3_pen45.shp", layer: "All_pRAN_VD_NS3_pen45"
## with 3041 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 3037 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 19%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 15 0 0 0 2 0 0 0 0 0 0 0.11764706
## 20 0 334 31 0 0 0 0 3 5 65 32 0.28936170
## 21 0 60 94 0 0 0 0 0 3 9 0 0.43373494
## 23 0 0 0 213 0 15 6 0 0 41 14 0.26297578
## 3 3 0 0 0 128 0 0 34 0 0 0 0.22424242
## 30 0 0 0 17 0 65 3 0 0 1 4 0.27777778
## 31 0 0 0 14 0 3 110 0 0 0 0 0.13385827
## 501 0 2 0 0 31 0 0 466 1 0 10 0.08627451
## 502 0 7 0 0 0 0 0 2 49 0 3 0.19672131
## 503 0 39 6 26 0 1 0 0 0 635 16 0.12171508
## 7 0 27 0 4 0 4 0 14 2 17 351 0.16229117map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")ns5
SNA VD ns5 15
mn="All_pSNA_VD_NS5_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_VD_NS5_pen15.shp", layer: "All_pSNA_VD_NS5_pen15"
## with 355 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 355 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 21.69%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 1 0 0 0 1 0 0 0 0 0 0 0.50000000
## 20 0 33 3 0 0 0 0 0 0 5 7 0.31250000
## 21 0 8 10 0 0 0 0 1 0 2 0 0.52380952
## 23 0 0 0 22 0 1 0 0 0 3 4 0.26666667
## 3 1 0 0 0 16 0 0 3 0 0 0 0.20000000
## 30 0 0 0 7 0 0 0 0 0 0 1 1.00000000
## 31 0 0 0 1 0 0 12 0 0 0 0 0.07692308
## 501 0 0 0 0 2 0 0 61 0 0 2 0.06153846
## 502 0 3 0 0 0 0 0 3 1 0 1 0.87500000
## 503 0 2 1 1 0 0 0 0 0 80 3 0.08045977
## 7 0 4 0 0 0 0 0 2 0 5 42 0.20754717map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA VD ns5 30
mn="All_pSNA_VD_NS5_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_VD_NS5_pen30.shp", layer: "All_pSNA_VD_NS5_pen30"
## with 675 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 674 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 17.36%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 2 0 0 0 1 0 0 0 0 0 0 0.33333333
## 20 0 65 5 0 0 0 0 1 0 11 6 0.26136364
## 21 0 11 20 0 0 0 0 0 0 1 0 0.37500000
## 23 0 0 0 44 0 5 1 0 0 2 3 0.20000000
## 3 1 0 0 0 35 0 0 3 0 0 0 0.10256410
## 30 0 0 0 12 0 2 0 0 0 0 1 0.86666667
## 31 0 0 0 1 0 0 25 0 0 0 0 0.03846154
## 501 0 0 0 0 5 0 0 123 1 0 1 0.05384615
## 502 0 3 0 0 0 0 0 3 7 0 2 0.53333333
## 503 0 9 0 2 0 0 0 0 0 152 5 0.09523810
## 7 0 8 0 1 0 0 0 2 0 10 82 0.20388350map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA VD ns5 45
mn="All_pSNA_VD_NS5_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_VD_NS5_pen45.shp", layer: "All_pSNA_VD_NS5_pen45"
## with 993 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 992 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 15.32%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 3 0 0 0 1 0 0 0 0 0 0 0.25000000
## 20 0 104 3 0 0 0 0 0 1 15 8 0.20610687
## 21 0 18 25 0 0 0 0 0 0 1 0 0.43181818
## 23 0 0 0 67 0 5 2 0 0 4 3 0.17283951
## 3 0 0 0 0 54 0 0 4 0 0 0 0.06896552
## 30 0 0 0 13 0 6 0 0 0 0 1 0.70000000
## 31 0 0 0 1 0 0 38 0 0 0 0 0.02564103
## 501 0 1 0 0 6 0 0 186 0 0 1 0.04123711
## 502 0 5 0 0 0 0 0 4 12 0 1 0.45454545
## 503 0 15 0 3 0 0 0 0 0 224 8 0.10400000
## 7 0 11 0 1 0 0 0 2 0 14 121 0.18791946map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN VD ns5 15
mn="All_pRAN_VD_NS5_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_VD_NS5_pen15.shp", layer: "All_pRAN_VD_NS5_pen15"
## with 355 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 355 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 32.68%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 1 0 0 0 1 0 0 0 0 0 0 0.5000000
## 20 0 22 8 0 0 0 0 0 0 10 8 0.5416667
## 21 0 13 6 0 0 0 0 0 0 1 1 0.7142857
## 23 0 0 0 17 0 3 4 0 0 5 1 0.4333333
## 3 0 0 0 0 12 0 0 8 0 0 0 0.4000000
## 30 0 0 0 7 0 1 0 0 0 0 0 0.8750000
## 31 0 0 0 5 0 0 8 0 0 0 0 0.3846154
## 501 0 0 0 0 7 0 0 57 0 0 1 0.1230769
## 502 0 1 0 0 0 0 0 3 1 0 3 0.8750000
## 503 0 7 0 3 0 0 0 0 0 74 3 0.1494253
## 7 0 7 0 1 0 0 0 2 0 3 40 0.2452830map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN VD ns5 30
mn="All_pRAN_VD_NS5_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_VD_NS5_pen30.shp", layer: "All_pRAN_VD_NS5_pen30"
## with 675 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 675 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 24.15%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 1 0 0 0 3 0 0 0 0 0 0 0.75000000
## 20 0 57 10 0 0 0 0 0 0 12 9 0.35227273
## 21 0 14 17 0 0 0 0 0 0 1 0 0.46875000
## 23 0 0 0 34 0 5 2 0 0 11 3 0.38181818
## 3 1 0 0 0 29 0 0 9 0 0 0 0.25641026
## 30 0 0 0 7 0 5 3 0 0 0 0 0.66666667
## 31 0 0 0 4 0 2 20 0 0 0 0 0.23076923
## 501 0 0 0 0 8 0 0 120 1 0 1 0.07692308
## 502 0 3 1 0 0 0 0 2 8 0 1 0.46666667
## 503 0 9 2 10 0 0 0 0 0 145 2 0.13690476
## 7 0 13 0 2 0 0 0 4 1 7 76 0.26213592map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN VD ns5 45
mn="All_pRAN_VD_NS5_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_VD_NS5_pen45.shp", layer: "All_pRAN_VD_NS5_pen45"
## with 993 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 992 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 17.04%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 3 0 0 0 1 0 0 0 0 0 0 0.25000000
## 20 0 93 9 0 0 0 0 1 0 15 13 0.29007634
## 21 0 15 27 0 0 0 0 0 0 2 0 0.38636364
## 23 0 0 0 65 0 3 3 0 0 9 1 0.19753086
## 3 0 0 0 0 44 0 0 14 0 0 0 0.24137931
## 30 0 0 0 9 0 8 3 0 0 0 0 0.60000000
## 31 0 0 0 3 0 1 33 0 0 0 2 0.15384615
## 501 0 0 0 0 7 0 0 185 0 0 2 0.04639175
## 502 0 1 0 0 0 0 0 3 17 0 1 0.22727273
## 503 0 10 1 7 0 0 0 0 0 229 3 0.08400000
## 7 0 14 0 1 0 0 1 3 1 10 119 0.20134228map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")ns7
SNA VD ns7 15
mn="All_pSNA_VD_NS7_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_VD_NS7_pen15.shp", layer: "All_pSNA_VD_NS7_pen15"
## with 257 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 256 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 13.28%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.00000000
## 20 0 29 2 0 0 0 0 0 0 4 0 0.17142857
## 21 0 2 6 0 0 0 0 0 0 0 0 0.25000000
## 23 0 0 0 22 0 0 0 0 0 0 0 0.00000000
## 3 0 0 0 0 12 0 0 2 0 0 0 0.14285714
## 30 0 0 0 2 0 2 0 0 0 0 0 0.50000000
## 31 0 0 0 3 0 0 5 0 0 0 0 0.37500000
## 501 0 0 0 0 0 0 0 51 0 0 1 0.01923077
## 502 0 2 0 0 0 0 0 4 0 0 1 1.00000000
## 503 0 2 0 0 0 0 0 0 0 57 3 0.08064516
## 7 0 3 0 0 0 0 0 0 0 2 38 0.11627907map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA VD ns7 30
mn="All_pSNA_VD_NS7_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_VD_NS7_pen30.shp", layer: "All_pSNA_VD_NS7_pen30"
## with 227 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 226 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 19.47%
## Confusion matrix:
## 20 21 23 3 30 31 501 502 503 7 class.error
## 20 16 0 0 0 0 0 0 0 5 6 0.4074074
## 21 3 2 0 0 0 0 0 0 0 0 0.6000000
## 23 0 0 15 0 1 1 0 0 1 1 0.2105263
## 3 0 0 0 13 0 0 0 0 0 0 0.0000000
## 30 0 0 1 0 0 1 0 0 0 1 1.0000000
## 31 0 0 2 0 0 5 0 0 0 0 0.2857143
## 501 0 0 0 0 0 0 51 0 0 0 0.0000000
## 502 2 0 0 0 0 0 3 0 0 1 1.0000000
## 503 2 0 0 0 0 0 0 0 52 3 0.0877193
## 7 2 0 1 0 0 0 1 0 6 28 0.2631579map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA VD ns7 45
mn="All_pSNA_VD_NS7_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_VD_NS7_pen45.shp", layer: "All_pSNA_VD_NS7_pen45"
## with 323 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 322 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 18.63%
## Confusion matrix:
## 20 21 23 3 30 31 501 502 503 7 class.error
## 20 23 1 0 0 0 0 0 0 6 5 0.34285714
## 21 4 4 0 0 0 0 0 0 0 0 0.50000000
## 23 0 0 21 0 2 2 0 0 1 2 0.25000000
## 3 0 0 0 18 0 0 2 0 0 0 0.10000000
## 30 0 0 3 0 0 0 0 0 0 1 1.00000000
## 31 0 0 2 0 0 8 0 0 0 0 0.20000000
## 501 0 0 0 1 0 0 73 0 0 1 0.02666667
## 502 2 0 0 0 0 0 4 0 0 1 1.00000000
## 503 4 0 1 0 0 0 0 0 72 4 0.11111111
## 7 2 0 1 0 0 0 1 0 7 43 0.20370370map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN VD ns7 15
mn="All_pRAN_VD_NS7_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_VD_NS7_pen15.shp", layer: "All_pRAN_VD_NS7_pen15"
## with 129 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 129 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 24.03%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.00000000
## 20 0 10 1 0 0 0 0 0 0 1 5 0.41176471
## 21 0 4 0 0 0 0 0 0 0 0 0 1.00000000
## 23 0 0 0 8 0 1 0 0 0 2 0 0.27272727
## 3 0 0 0 0 7 0 0 0 0 0 0 0.00000000
## 30 0 0 0 2 0 0 0 0 0 0 0 1.00000000
## 31 0 0 0 0 0 0 4 0 0 0 0 0.00000000
## 501 0 0 0 0 0 0 0 25 1 0 0 0.03846154
## 502 0 2 0 0 0 0 0 1 0 0 1 1.00000000
## 503 0 1 0 2 0 0 0 0 0 26 2 0.16129032
## 7 0 2 0 0 0 0 0 1 0 1 18 0.18181818map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN VD ns7 30
mn="All_pRAN_VD_NS7_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_VD_NS7_pen30.shp", layer: "All_pRAN_VD_NS7_pen30"
## with 227 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 227 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 22.47%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.00000000
## 20 0 17 0 0 0 0 0 0 0 5 5 0.37037037
## 21 0 3 2 0 0 0 0 0 0 0 0 0.60000000
## 23 0 0 0 12 0 3 0 0 0 4 0 0.36842105
## 3 1 0 0 0 11 0 0 1 0 0 0 0.15384615
## 30 0 0 0 3 0 0 0 0 0 0 0 1.00000000
## 31 0 0 0 2 0 0 5 0 0 0 0 0.28571429
## 501 0 0 0 0 1 0 0 48 0 0 2 0.05882353
## 502 0 1 0 0 0 0 0 2 3 0 0 0.50000000
## 503 0 4 0 3 0 0 0 0 0 48 2 0.15789474
## 7 0 2 0 1 0 0 0 1 0 4 30 0.21052632map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN VD ns7 45
mn="All_pRAN_VD_NS7_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_VD_NS7_pen45.shp", layer: "All_pRAN_VD_NS7_pen45"
## with 323 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 323 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 16.72%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.00000000
## 20 0 26 1 0 0 0 0 0 1 2 5 0.25714286
## 21 0 4 4 0 0 0 0 0 0 0 0 0.50000000
## 23 0 0 0 19 0 2 0 0 0 6 1 0.32142857
## 3 0 0 0 0 18 0 0 2 0 0 0 0.10000000
## 30 0 0 0 4 0 0 0 0 0 0 0 1.00000000
## 31 0 0 0 2 0 0 8 0 0 0 0 0.20000000
## 501 0 0 0 0 3 0 0 71 0 0 1 0.05333333
## 502 0 4 0 0 0 0 0 3 0 0 0 1.00000000
## 503 0 0 0 3 0 0 0 0 0 78 0 0.03703704
## 7 0 3 0 1 0 0 0 1 0 4 45 0.16666667map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RSP -ok
ns3
SNA RSP ns3 15
mn="All_pSNA_RSP_NS3_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_RSP_NS3_pen15.shp", layer: "All_pSNA_RSP_NS3_pen15"
## with 1045 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 1043 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 18.31%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 3 0 0 0 3 0 0 0 0 0 0 0.50000000
## 20 0 125 7 0 0 0 0 1 0 23 9 0.24242424
## 21 0 25 35 0 0 0 0 0 0 2 0 0.43548387
## 23 0 0 0 83 0 5 0 0 0 12 2 0.18627451
## 3 1 0 0 0 44 0 0 10 0 0 0 0.20000000
## 30 0 0 0 11 0 22 0 0 0 0 0 0.33333333
## 31 0 0 0 2 0 3 35 0 0 0 0 0.12500000
## 501 0 0 0 0 8 0 0 159 1 0 2 0.06470588
## 502 0 0 1 0 0 0 0 3 16 1 1 0.27272727
## 503 0 16 0 4 0 0 0 0 0 211 13 0.13524590
## 7 0 10 0 1 0 0 0 3 0 11 119 0.17361111map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA RSP ns3 30
mn="All_pSNA_RSP_NS3_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_RSP_NS3_pen30.shp", layer: "All_pSNA_RSP_NS3_pen30"
## with 2037 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 2035 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 15.04%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 8 0 0 0 4 0 0 0 0 0 0 0.33333333
## 20 0 246 13 0 0 0 0 2 0 41 16 0.22641509
## 21 0 33 76 0 0 0 0 0 1 4 0 0.33333333
## 23 0 0 0 170 0 8 0 0 0 12 5 0.12820513
## 3 3 0 0 0 96 0 0 11 0 0 0 0.12727273
## 30 0 0 0 16 0 45 0 0 0 1 0 0.27419355
## 31 0 0 0 4 0 2 74 0 0 0 0 0.07500000
## 501 0 0 0 0 12 0 0 320 2 0 6 0.05882353
## 502 0 5 2 0 0 0 0 4 27 0 3 0.34146341
## 503 0 26 0 11 0 0 0 1 0 427 17 0.11410788
## 7 0 20 0 2 0 0 0 6 1 12 240 0.14590747map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA RSP ns3 45
mn="All_pSNA_RSP_NS3_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_RSP_NS3_pen45.shp", layer: "All_pSNA_RSP_NS3_pen45"
## with 3037 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 3032 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 15.11%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 12 0 0 0 3 0 0 0 0 0 0 0.20000000
## 20 0 370 19 0 0 0 0 1 1 49 31 0.21443737
## 21 0 59 104 0 0 0 0 0 0 7 0 0.38823529
## 23 0 0 0 234 0 16 2 0 0 24 12 0.18750000
## 3 3 0 0 0 141 0 0 21 0 0 0 0.14545455
## 30 0 0 0 13 0 73 2 0 0 1 2 0.19780220
## 31 0 0 0 4 0 4 112 0 0 0 0 0.06666667
## 501 0 0 0 0 15 0 0 486 1 0 8 0.04705882
## 502 0 4 0 0 0 0 0 7 44 0 5 0.26666667
## 503 0 41 0 18 0 1 0 2 0 640 20 0.11357341
## 7 0 28 0 3 0 0 0 9 2 20 358 0.14761905map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN RSP ns3 15
mn="All_pRAN_RSP_NS3_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_RSP_NS3_pen15.shp", layer: "All_pRAN_RSP_NS3_pen15"
## with 1045 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 1043 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 24.64%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 4 0 0 0 2 0 0 0 0 0 0 0.3333333
## 20 0 111 13 1 0 0 0 0 1 23 16 0.3272727
## 21 0 27 28 0 0 0 0 0 1 5 1 0.5483871
## 23 0 0 0 74 0 7 1 0 0 18 2 0.2745098
## 3 1 0 0 0 43 0 0 11 0 0 0 0.2181818
## 30 0 0 0 6 0 22 3 0 0 1 1 0.3333333
## 31 0 0 0 6 0 4 29 0 0 1 0 0.2750000
## 501 0 1 0 0 13 0 0 151 0 0 5 0.1117647
## 502 0 7 1 0 0 0 0 2 10 0 2 0.5454545
## 503 0 22 1 10 0 1 0 0 0 204 6 0.1639344
## 7 0 16 0 1 0 1 0 8 1 7 110 0.2361111map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN RSP ns3 30
mn="All_pRAN_RSP_NS3_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_RSP_NS3_pen30.shp", layer: "All_pRAN_RSP_NS3_pen30"
## with 2037 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 2034 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 20.94%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 10 0 0 0 1 0 0 0 0 0 0 0.09090909
## 20 0 227 18 0 0 0 0 6 0 46 21 0.28616352
## 21 0 48 58 0 0 0 0 1 1 6 0 0.49122807
## 23 0 0 0 148 0 10 6 0 0 23 8 0.24102564
## 3 1 0 0 0 84 0 0 25 0 0 0 0.23636364
## 30 0 0 0 14 0 43 2 0 0 2 1 0.30645161
## 31 0 0 0 15 0 2 63 0 0 0 0 0.21250000
## 501 0 3 0 0 20 0 0 306 1 0 10 0.10000000
## 502 0 3 2 0 1 0 0 1 30 1 3 0.26829268
## 503 0 30 2 15 0 0 1 0 1 417 16 0.13485477
## 7 0 21 2 3 0 3 0 15 2 13 222 0.20996441map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN RSP ns3 45
mn="All_pRAN_RSP_NS3_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_RSP_NS3_pen45.shp", layer: "All_pRAN_RSP_NS3_pen45"
## with 3037 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 3034 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 18.19%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 15 0 0 0 2 0 0 0 0 0 0 0.11764706
## 20 0 353 18 0 0 0 0 4 1 64 31 0.25053079
## 21 0 58 103 0 0 0 0 0 0 7 2 0.39411765
## 23 0 0 0 221 0 18 6 0 0 32 11 0.23263889
## 3 2 0 0 0 131 0 0 32 0 0 0 0.20606061
## 30 0 0 0 15 0 66 5 0 0 2 3 0.27472527
## 31 0 0 0 16 0 5 99 0 0 0 0 0.17500000
## 501 0 3 0 0 22 0 0 471 1 0 13 0.07647059
## 502 0 7 1 0 0 0 0 1 44 1 6 0.26666667
## 503 0 43 2 20 0 2 0 0 1 638 16 0.11634349
## 7 0 27 2 2 0 2 0 20 4 22 341 0.18809524map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")ns5
SNA RSP ns5 15
mn="All_pSNA_RSP_NS5_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_RSP_NS5_pen15.shp", layer: "All_pSNA_RSP_NS5_pen15"
## with 355 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 355 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 21.97%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 1 0 0 0 1 0 0 0 0 0 0 0.50000000
## 20 0 31 6 0 0 0 0 0 0 8 3 0.35416667
## 21 0 8 10 0 0 0 0 1 0 3 0 0.54545455
## 23 0 0 0 27 0 2 0 0 0 0 1 0.10000000
## 3 1 0 0 0 16 0 0 3 0 0 0 0.20000000
## 30 0 0 0 4 0 4 0 0 0 0 0 0.50000000
## 31 0 0 0 0 0 1 12 0 0 0 0 0.07692308
## 501 0 0 0 0 2 0 0 61 0 0 2 0.06153846
## 502 0 3 0 0 0 0 0 2 2 0 0 0.71428571
## 503 0 3 1 0 0 0 0 0 0 76 7 0.12643678
## 7 0 4 0 0 0 0 0 2 0 10 37 0.30188679map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA RSP ns5 30
mn="All_pSNA_RSP_NS5_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_RSP_NS5_pen30.shp", layer: "All_pSNA_RSP_NS5_pen30"
## with 674 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 673 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 16.79%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 2 0 0 0 1 0 0 0 0 0 0 0.33333333
## 20 0 66 3 0 0 0 0 1 0 12 8 0.26666667
## 21 0 12 20 0 0 0 0 0 0 1 0 0.39393939
## 23 0 0 0 46 0 3 2 0 0 0 4 0.16363636
## 3 1 0 0 0 35 0 0 3 0 0 0 0.10256410
## 30 0 0 0 8 0 6 0 0 0 0 1 0.60000000
## 31 0 0 0 0 0 1 25 0 0 0 0 0.03846154
## 501 0 0 0 0 4 0 0 124 0 0 1 0.03875969
## 502 0 3 1 0 0 0 0 3 5 0 2 0.64285714
## 503 0 9 0 1 0 0 0 0 0 152 6 0.09523810
## 7 0 6 0 2 0 0 0 2 0 12 79 0.21782178map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA RSP ns5 45
mn="All_pSNA_RSP_NS5_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_RSP_NS5_pen45.shp", layer: "All_pSNA_RSP_NS5_pen45"
## with 992 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 991 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 15.44%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 3 0 0 0 1 0 0 0 0 0 0 0.25000000
## 20 0 97 7 0 0 0 0 1 0 16 9 0.25384615
## 21 0 14 30 0 0 0 0 0 0 1 0 0.33333333
## 23 0 0 0 70 0 4 2 0 0 2 3 0.13580247
## 3 1 0 0 0 52 0 0 5 0 0 0 0.10344828
## 30 0 0 0 8 0 10 1 0 0 0 1 0.50000000
## 31 0 0 0 1 0 2 36 0 0 0 0 0.07692308
## 501 0 0 0 0 6 0 0 184 0 0 2 0.04166667
## 502 0 6 1 0 0 0 0 6 7 0 1 0.66666667
## 503 0 11 0 4 0 0 0 0 0 230 7 0.08730159
## 7 0 10 0 2 0 0 0 5 0 13 119 0.20134228map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN RSP ns5 15
mn="All_pRAN_RSP_NS5_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_RSP_NS5_pen15.shp", layer: "All_pRAN_RSP_NS5_pen15"
## with 355 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 355 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 25.63%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 2 0 0 0 0 0 0 0 0 0 0 0.0000000
## 20 0 33 5 0 0 0 0 1 0 6 3 0.3125000
## 21 0 11 10 0 0 0 0 0 0 1 0 0.5454545
## 23 0 0 0 18 0 2 3 0 0 5 2 0.4000000
## 3 0 0 0 0 13 0 0 7 0 0 0 0.3500000
## 30 0 0 0 5 0 1 2 0 0 0 0 0.8750000
## 31 0 0 0 3 0 0 10 0 0 0 0 0.2307692
## 501 0 0 0 0 7 0 0 58 0 0 0 0.1076923
## 502 0 0 0 0 0 0 0 1 5 0 1 0.2857143
## 503 0 6 2 1 0 0 0 0 0 76 2 0.1264368
## 7 0 9 0 3 0 0 0 2 0 1 38 0.2830189map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN RSP ns5 30
mn="All_pRAN_RSP_NS5_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_RSP_NS5_pen30.shp", layer: "All_pRAN_RSP_NS5_pen30"
## with 674 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 674 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 22.26%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 2 0 0 0 2 0 0 0 0 0 0 0.5000000
## 20 0 63 8 0 0 0 0 0 2 11 6 0.3000000
## 21 0 15 16 0 0 0 0 0 0 2 0 0.5151515
## 23 0 0 0 44 0 4 0 0 0 6 1 0.2000000
## 3 1 0 0 0 25 0 0 13 0 0 0 0.3589744
## 30 0 0 0 5 0 7 2 0 0 0 1 0.5333333
## 31 0 0 0 1 0 3 22 0 0 0 0 0.1538462
## 501 0 1 0 0 11 0 0 115 0 0 2 0.1085271
## 502 0 3 1 0 0 0 0 4 4 1 1 0.7142857
## 503 0 8 3 8 0 0 0 0 0 144 5 0.1428571
## 7 0 7 0 1 0 1 0 2 0 8 82 0.1881188map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN RSP ns5 45
mn="All_pRAN_RSP_NS5_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_RSP_NS5_pen45.shp", layer: "All_pRAN_RSP_NS5_pen45"
## with 992 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 992 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 17.84%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 3 0 0 0 2 0 0 0 0 0 0 0.40000000
## 20 0 93 8 0 0 0 0 0 1 15 13 0.28461538
## 21 0 16 28 0 0 0 0 0 0 1 0 0.37777778
## 23 0 0 0 57 0 5 2 0 0 13 4 0.29629630
## 3 0 0 0 0 46 0 0 12 0 0 0 0.20689655
## 30 0 0 0 8 0 7 5 0 0 0 0 0.65000000
## 31 0 0 0 2 0 1 36 0 0 0 0 0.07692308
## 501 0 0 0 0 9 0 0 181 0 0 2 0.05729167
## 502 0 3 0 0 0 0 0 2 15 0 1 0.28571429
## 503 0 10 2 6 0 0 0 0 0 228 6 0.09523810
## 7 0 11 1 2 0 0 0 8 0 6 121 0.18791946map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")ns7
SNA RSP ns7 15
mn="All_pSNA_RSP_NS7_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_RSP_NS7_pen15.shp", layer: "All_pSNA_RSP_NS7_pen15"
## with 127 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 126 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 20.63%
## Confusion matrix:
## 20 21 23 3 30 31 501 502 503 7 class.error
## 20 12 1 0 0 0 0 0 0 4 0 0.2941176
## 21 2 2 0 0 0 0 0 0 0 0 0.5000000
## 23 0 0 11 0 0 0 0 0 0 0 0.0000000
## 3 0 0 0 6 0 0 1 0 0 0 0.1428571
## 30 0 0 1 0 0 0 0 0 0 1 1.0000000
## 31 0 0 1 0 0 3 0 0 0 0 0.2500000
## 501 0 0 0 0 0 0 26 0 0 0 0.0000000
## 502 1 0 0 0 0 0 2 0 0 0 1.0000000
## 503 2 0 0 0 0 0 0 0 26 3 0.1612903
## 7 1 0 0 0 0 0 0 0 6 14 0.3333333map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA RSP ns7 30
mn="All_pSNA_RSP_NS7_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_RSP_NS7_pen30.shp", layer: "All_pSNA_RSP_NS7_pen30"
## with 226 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 225 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 18.67%
## Confusion matrix:
## 20 21 23 3 30 31 501 502 503 7 class.error
## 20 18 0 0 0 0 0 0 0 7 2 0.3333333
## 21 3 2 0 0 0 0 1 0 0 0 0.6666667
## 23 0 0 18 0 0 0 0 0 1 1 0.1000000
## 3 0 0 0 13 0 0 0 0 0 0 0.0000000
## 30 0 0 2 0 0 0 0 0 0 1 1.0000000
## 31 0 0 0 0 0 7 0 0 0 0 0.0000000
## 501 0 0 0 0 0 0 49 0 0 1 0.0200000
## 502 1 0 0 0 0 0 4 0 0 0 1.0000000
## 503 5 0 0 0 0 0 0 0 50 2 0.1228070
## 7 2 0 0 0 0 0 1 0 8 26 0.2972973map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")SNA RSP ns7 45
mn="All_pSNA_RSP_NS7_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pSNA_RSP_NS7_pen45.shp", layer: "All_pSNA_RSP_NS7_pen45"
## with 321 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 320 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 15%
## Confusion matrix:
## 20 21 23 3 30 31 501 502 503 7 class.error
## 20 27 1 0 0 0 0 0 0 4 4 0.25000000
## 21 4 4 0 0 0 0 0 0 0 0 0.50000000
## 23 0 0 24 0 1 1 0 0 0 2 0.14285714
## 3 0 0 0 18 0 0 1 0 0 0 0.05263158
## 30 0 0 3 0 0 0 0 0 0 1 1.00000000
## 31 0 0 2 0 0 8 0 0 0 0 0.20000000
## 501 0 0 0 1 0 0 73 0 0 1 0.02666667
## 502 3 0 0 0 0 0 3 0 0 0 1.00000000
## 503 3 0 0 0 0 0 0 0 77 3 0.07228916
## 7 2 0 0 0 0 0 1 0 7 41 0.19607843map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN RSP ns7 15
mn="All_pRAN_RSP_NS7_pen15"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_RSP_NS7_pen15.shp", layer: "All_pRAN_RSP_NS7_pen15"
## with 127 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 127 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 25.98%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.00000000
## 20 0 11 1 0 0 0 0 0 0 2 3 0.35294118
## 21 0 3 1 0 0 0 0 0 0 0 0 0.75000000
## 23 0 0 0 7 0 0 2 0 0 2 0 0.36363636
## 3 0 0 0 0 7 0 0 0 0 0 0 0.00000000
## 30 0 0 0 1 0 0 0 0 0 0 1 1.00000000
## 31 0 0 0 1 0 0 3 0 0 0 0 0.25000000
## 501 0 0 0 0 1 0 0 24 1 0 0 0.07692308
## 502 0 2 0 0 0 0 0 1 0 0 0 1.00000000
## 503 0 3 0 1 0 0 0 0 0 26 1 0.16129032
## 7 0 2 0 0 0 1 0 0 0 3 15 0.28571429map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN RSP ns7 30
mn="All_pRAN_RSP_NS7_pen30"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_RSP_NS7_pen30.shp", layer: "All_pRAN_RSP_NS7_pen30"
## with 226 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 226 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 22.57%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.0000000
## 20 0 16 1 0 0 0 0 0 0 5 5 0.4074074
## 21 0 4 1 0 0 0 0 1 0 0 0 0.8333333
## 23 0 0 0 17 0 0 1 0 0 2 0 0.1500000
## 3 0 0 0 0 11 0 0 2 0 0 0 0.1538462
## 30 0 0 0 1 0 1 1 0 0 0 0 0.6666667
## 31 0 0 0 1 0 1 5 0 0 0 0 0.2857143
## 501 0 0 0 0 2 0 0 48 0 0 0 0.0400000
## 502 0 1 0 0 0 0 0 3 1 0 0 0.8000000
## 503 0 4 0 2 0 0 0 0 0 48 3 0.1578947
## 7 0 3 0 1 0 0 0 1 0 5 27 0.2702703map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")RAN RSP ns7 45
mn="All_pRAN_RSP_NS7_pen45"
shpPoint<-readOGR(paste(point_shp_folder,mn,".shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\allTrainingSamples_shp\All_pRAN_RSP_NS7_pen45.shp", layer: "All_pRAN_RSP_NS7_pen45"
## with 321 features
## It has 1 fieldssampData <- extract(covStack, shpPoint, sp = TRUE,method = "simple")
sampData<-sampData@data
sampData<-sampData[sampData$soilID %in% soil_type,]
sampData<-na.omit(sampData)
print(dim(sampData))## [1] 321 10sampleNum_dict[mn]<-nrow(sampData)
sampData$Bedrock<- droplevels(as.factor(sampData$Bedrock))
sampData$soilID<- droplevels(as.factor(sampData$soilID))
RF_model<-randomForest(soilID ~.,data = sampData,
importance = TRUE, proximity = FALSE,
ntree=1000,type="classification")
varImpPlot(RF_model)
print(RF_model)##
## Call:
## randomForest(formula = soilID ~ ., data = sampData, importance = TRUE, proximity = FALSE, ntree = 1000, type = "classification")
## Type of random forest: classification
## Number of trees: 1000
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 17.13%
## Confusion matrix:
## 1 20 21 23 3 30 31 501 502 503 7 class.error
## 1 0 0 0 0 1 0 0 0 0 0 0 1.00000000
## 20 0 25 1 0 0 0 0 0 0 4 6 0.30555556
## 21 0 4 4 0 0 0 0 0 0 0 0 0.50000000
## 23 0 0 0 20 0 1 2 0 0 5 0 0.28571429
## 3 1 0 0 0 16 0 0 2 0 0 0 0.15789474
## 30 0 0 0 1 0 3 0 0 0 0 0 0.25000000
## 31 0 0 0 1 0 0 9 0 0 0 0 0.10000000
## 501 0 0 0 0 1 0 0 73 0 0 1 0.02666667
## 502 0 1 0 0 0 0 0 3 1 0 1 0.83333333
## 503 0 4 0 5 0 0 0 0 0 72 2 0.13253012
## 7 0 3 0 0 0 0 0 1 0 4 43 0.15686275map_RF <- predict(covStack, RF_model, paste(result_tif_folder,mn,".tif",sep=""),
format = "GTiff", datatype = "FLT4S", overwrite = TRUE)
rm("mn","shpPoint","sampData","RF_model","map_RF")计算精度
#设置最终结果tif的保存路径
result_tif_folder="E:/soilPaper/modling/result_mapTif/"
result_map_name<-c("All_pSNA_DEM_NS3_pen15","All_pSNA_DEM_NS5_pen15","All_pSNA_DEM_NS7_pen15",
"All_pRAN_DEM_NS3_pen15","All_pRAN_DEM_NS5_pen15","All_pRAN_DEM_NS7_pen15",
"All_pSNA_DEM_NS3_pen30","All_pSNA_DEM_NS5_pen30","All_pSNA_DEM_NS7_pen30",
"All_pRAN_DEM_NS3_pen30","All_pRAN_DEM_NS5_pen30","All_pRAN_DEM_NS7_pen30",
"All_pSNA_DEM_NS3_pen45","All_pSNA_DEM_NS5_pen45","All_pSNA_DEM_NS7_pen45",
"All_pRAN_DEM_NS3_pen45","All_pRAN_DEM_NS5_pen45","All_pRAN_DEM_NS7_pen45",
"All_pSNA_Slope_NS3_pen15","All_pSNA_Slope_NS5_pen15","All_pSNA_Slope_NS7_pen15",
"All_pRAN_Slope_NS3_pen15","All_pRAN_Slope_NS5_pen15","All_pRAN_Slope_NS7_pen15",
"All_pSNA_Slope_NS3_pen30","All_pSNA_Slope_NS5_pen30","All_pSNA_Slope_NS7_pen30",
"All_pRAN_Slope_NS3_pen30","All_pRAN_Slope_NS5_pen30","All_pRAN_Slope_NS7_pen30",
"All_pSNA_Slope_NS3_pen45","All_pSNA_Slope_NS5_pen45","All_pSNA_Slope_NS7_pen45",
"All_pRAN_Slope_NS3_pen45","All_pRAN_Slope_NS5_pen45","All_pRAN_Slope_NS7_pen45",
"All_pSNA_PlanC_NS3_pen15","All_pSNA_PlanC_NS5_pen15","All_pSNA_PlanC_NS7_pen15",
"All_pRAN_PlanC_NS3_pen15","All_pRAN_PlanC_NS5_pen15","All_pRAN_PlanC_NS7_pen15",
"All_pSNA_PlanC_NS3_pen30","All_pSNA_PlanC_NS5_pen30","All_pSNA_PlanC_NS7_pen30",
"All_pRAN_PlanC_NS3_pen30","All_pRAN_PlanC_NS5_pen30","All_pRAN_PlanC_NS7_pen30",
"All_pSNA_PlanC_NS3_pen45","All_pSNA_PlanC_NS5_pen45","All_pSNA_PlanC_NS7_pen45",
"All_pRAN_PlanC_NS3_pen45","All_pRAN_PlanC_NS5_pen45","All_pRAN_PlanC_NS7_pen45",
"All_pSNA_ProfileC_NS3_pen15","All_pSNA_ProfileC_NS5_pen15","All_pSNA_ProfileC_NS7_pen15",
"All_pRAN_ProfileC_NS3_pen15","All_pRAN_ProfileC_NS5_pen15","All_pRAN_ProfileC_NS7_pen15",
"All_pSNA_ProfileC_NS3_pen30","All_pSNA_ProfileC_NS5_pen30","All_pSNA_ProfileC_NS7_pen30",
"All_pRAN_ProfileC_NS3_pen30","All_pRAN_ProfileC_NS5_pen30","All_pRAN_ProfileC_NS7_pen30",
"All_pSNA_ProfileC_NS3_pen45","All_pSNA_ProfileC_NS5_pen45","All_pSNA_ProfileC_NS7_pen45",
"All_pRAN_ProfileC_NS3_pen45","All_pRAN_ProfileC_NS5_pen45","All_pRAN_ProfileC_NS7_pen45",
"All_pSNA_TWI_NS3_pen15","All_pSNA_TWI_NS5_pen15","All_pSNA_TWI_NS7_pen15",
"All_pRAN_TWI_NS3_pen15","All_pRAN_TWI_NS5_pen15","All_pRAN_TWI_NS7_pen15",
"All_pSNA_TWI_NS3_pen30","All_pSNA_TWI_NS5_pen30","All_pSNA_TWI_NS7_pen30",
"All_pRAN_TWI_NS3_pen30","All_pRAN_TWI_NS5_pen30","All_pRAN_TWI_NS7_pen30",
"All_pSNA_TWI_NS3_pen45","All_pSNA_TWI_NS5_pen45","All_pSNA_TWI_NS7_pen45",
"All_pRAN_TWI_NS3_pen45","All_pRAN_TWI_NS5_pen45","All_pRAN_TWI_NS7_pen45",
"All_pSNA_LSF_NS3_pen15","All_pSNA_LSF_NS5_pen15","All_pSNA_LSF_NS7_pen15",
"All_pRAN_LSF_NS3_pen15","All_pRAN_LSF_NS5_pen15","All_pRAN_LSF_NS7_pen15",
"All_pSNA_LSF_NS3_pen30","All_pSNA_LSF_NS5_pen30","All_pSNA_LSF_NS7_pen30",
"All_pRAN_LSF_NS3_pen30","All_pRAN_LSF_NS5_pen30","All_pRAN_LSF_NS7_pen30",
"All_pSNA_LSF_NS3_pen45","All_pSNA_LSF_NS5_pen45","All_pSNA_LSF_NS7_pen45",
"All_pRAN_LSF_NS3_pen45","All_pRAN_LSF_NS5_pen45","All_pRAN_LSF_NS7_pen45",
"All_pSNA_VD_NS3_pen15","All_pSNA_VD_NS5_pen15","All_pSNA_VD_NS7_pen15",
"All_pRAN_VD_NS3_pen15","All_pRAN_VD_NS5_pen15","All_pRAN_VD_NS7_pen15",
"All_pSNA_VD_NS3_pen30","All_pSNA_VD_NS5_pen30","All_pSNA_VD_NS7_pen30",
"All_pRAN_VD_NS3_pen30","All_pRAN_VD_NS5_pen30","All_pRAN_VD_NS7_pen30",
"All_pSNA_VD_NS3_pen45","All_pSNA_VD_NS5_pen45","All_pSNA_VD_NS7_pen45",
"All_pRAN_VD_NS3_pen45","All_pRAN_VD_NS5_pen45","All_pRAN_VD_NS7_pen45",
"All_pSNA_RSP_NS3_pen15","All_pSNA_RSP_NS5_pen15","All_pSNA_RSP_NS7_pen15",
"All_pRAN_RSP_NS3_pen15","All_pRAN_RSP_NS5_pen15","All_pRAN_RSP_NS7_pen15",
"All_pSNA_RSP_NS3_pen30","All_pSNA_RSP_NS5_pen30","All_pSNA_RSP_NS7_pen30",
"All_pRAN_RSP_NS3_pen30","All_pRAN_RSP_NS5_pen30","All_pRAN_RSP_NS7_pen30",
"All_pSNA_RSP_NS3_pen45","All_pSNA_RSP_NS5_pen45","All_pSNA_RSP_NS7_pen45",
"All_pRAN_RSP_NS3_pen45","All_pRAN_RSP_NS5_pen45","All_pRAN_RSP_NS7_pen45")
resultCovStack<-raster(paste(result_tif_folder,result_map_name[1],".tif",sep = ""))
for (i in 2:length(result_map_name)) {
resultCovStack <- stack(resultCovStack, raster(paste(result_tif_folder,result_map_name[i],".tif",sep = "")))
}
validationPoint<-readOGR(paste(result_tif_folder,"validationPoint.shp",sep = "")) ##### OGR data source with driver: ESRI Shapefile
## Source: "E:\soilPaper\modling\result_mapTif\validationPoint.shp", layer: "validationPoint"
## with 99 features
## It has 3 fieldsvp_data <- extract(resultCovStack, validationPoint, sp = TRUE,method = "simple")
vp_data<-na.omit(vp_data@data)
vp_data<-vp_data[vp_data$trueType %in% soil_type,]
print(nrow(vp_data))## [1] 90precision_df<-data.frame("method"=rep(NA,length(result_map_name)),
"factor"=rep(NA,length(result_map_name)),
"NS"=rep(NA,length(result_map_name)),
"percent"=rep(NA,length(result_map_name)),
"precisionValue"=rep(NA,length(result_map_name)),
"sampleNum"=rep(NA,length(result_map_name)))
for (i in 1:length(result_map_name)){
fna<-result_map_name[i]
ns_split<-unlist(strsplit(fna,"_"))
if (length(ns_split)==5){
precision_df[i,]$method<-ns_split[2]
precision_df[i,]$factor<-ns_split[3]
precision_df[i,]$NS<-ns_split[4]
precision_df[i,]$percent<-ns_split[5]
precision_df[i,]$precisionValue<-round( sum(vp_data$trueType==vp_data[fna])/nrow(vp_data),4)
precision_df[i,]$sampleNum<-as.vector(unlist(sampleNum_dict[fna]))
}
}
precision_oldSoilMap<-round( sum(vp_data$trueType==vp_data$oldSoilMap)/nrow(vp_data),4)
print(precision_oldSoilMap)## [1] 0.6333print(precision_df)## method factor NS percent precisionValue sampleNum
## 1 pSNA DEM NS3 pen15 0.3222 1040
## 2 pSNA DEM NS5 pen15 0.2556 354
## 3 pSNA DEM NS7 pen15 0.4778 130
## 4 pRAN DEM NS3 pen15 0.6778 1042
## 5 pRAN DEM NS5 pen15 0.7000 356
## 6 pRAN DEM NS7 pen15 0.6778 131
## 7 pSNA DEM NS3 pen30 0.3778 2040
## 8 pSNA DEM NS5 pen30 0.4222 677
## 9 pSNA DEM NS7 pen30 0.3889 226
## 10 pRAN DEM NS3 pen30 0.6778 2042
## 11 pRAN DEM NS5 pen30 0.7000 679
## 12 pRAN DEM NS7 pen30 0.6889 227
## 13 pSNA DEM NS3 pen45 0.4667 3036
## 14 pSNA DEM NS5 pen45 0.4333 993
## 15 pSNA DEM NS7 pen45 0.3889 326
## 16 pRAN DEM NS3 pen45 0.6444 3038
## 17 pRAN DEM NS5 pen45 0.6778 995
## 18 pRAN DEM NS7 pen45 0.6778 327
## 19 pSNA Slope NS3 pen15 0.7333 994
## 20 pSNA Slope NS5 pen15 0.6667 336
## 21 pSNA Slope NS7 pen15 0.6667 121
## 22 pRAN Slope NS3 pen15 0.6889 996
## 23 pRAN Slope NS5 pen15 0.6000 336
## 24 pRAN Slope NS7 pen15 0.6444 121
## 25 pSNA Slope NS3 pen30 0.7000 1947
## 26 pSNA Slope NS5 pen30 0.6889 636
## 27 pSNA Slope NS7 pen30 0.6000 214
## 28 pRAN Slope NS3 pen30 0.6222 1948
## 29 pRAN Slope NS5 pen30 0.6333 636
## 30 pRAN Slope NS7 pen30 0.6444 214
## 31 pSNA Slope NS3 pen45 0.7333 2885
## 32 pSNA Slope NS5 pen45 0.6444 929
## 33 pSNA Slope NS7 pen45 0.6556 307
## 34 pRAN Slope NS3 pen45 0.6556 2886
## 35 pRAN Slope NS5 pen45 0.6333 931
## 36 pRAN Slope NS7 pen45 0.7000 307
## 37 pSNA PlanC NS3 pen15 0.6333 1011
## 38 pSNA PlanC NS5 pen15 0.6778 345
## 39 pSNA PlanC NS7 pen15 0.5333 129
## 40 pRAN PlanC NS3 pen15 0.6889 1011
## 41 pRAN PlanC NS5 pen15 0.5889 345
## 42 pRAN PlanC NS7 pen15 0.6556 129
## 43 pSNA PlanC NS3 pen30 0.6222 1973
## 44 pSNA PlanC NS5 pen30 0.6778 649
## 45 pSNA PlanC NS7 pen30 0.5222 227
## 46 pRAN PlanC NS3 pen30 0.6556 1973
## 47 pRAN PlanC NS5 pen30 0.6556 650
## 48 pRAN PlanC NS7 pen30 0.6222 227
## 49 pSNA PlanC NS3 pen45 0.6222 2937
## 50 pSNA PlanC NS5 pen45 0.6556 953
## 51 pSNA PlanC NS7 pen45 0.5667 319
## 52 pRAN PlanC NS3 pen45 0.6667 2937
## 53 pRAN PlanC NS5 pen45 0.6667 954
## 54 pRAN PlanC NS7 pen45 0.6889 320
## 55 pSNA ProfileC NS3 pen15 0.6444 996
## 56 pSNA ProfileC NS5 pen15 0.5778 338
## 57 pSNA ProfileC NS7 pen15 0.6222 122
## 58 pRAN ProfileC NS3 pen15 0.6889 998
## 59 pRAN ProfileC NS5 pen15 0.6556 340
## 60 pRAN ProfileC NS7 pen15 0.6778 122
## 61 pSNA ProfileC NS3 pen30 0.6444 1946
## 62 pSNA ProfileC NS5 pen30 0.7000 636
## 63 pSNA ProfileC NS7 pen30 0.5778 215
## 64 pRAN ProfileC NS3 pen30 0.6778 1946
## 65 pRAN ProfileC NS5 pen30 0.7111 638
## 66 pRAN ProfileC NS7 pen30 0.6667 215
## 67 pSNA ProfileC NS3 pen45 0.6444 2895
## 68 pSNA ProfileC NS5 pen45 0.6444 934
## 69 pSNA ProfileC NS7 pen45 0.6333 307
## 70 pRAN ProfileC NS3 pen45 0.6222 2894
## 71 pRAN ProfileC NS5 pen45 0.6667 936
## 72 pRAN ProfileC NS7 pen45 0.6778 307
## 73 pSNA TWI NS3 pen15 0.7333 983
## 74 pSNA TWI NS5 pen15 0.8000 337
## 75 pSNA TWI NS7 pen15 0.7111 122
## 76 pRAN TWI NS3 pen15 0.6667 983
## 77 pRAN TWI NS5 pen15 0.6444 337
## 78 pRAN TWI NS7 pen15 0.6333 122
## 79 pSNA TWI NS3 pen30 0.6889 1919
## 80 pSNA TWI NS5 pen30 0.7667 634
## 81 pSNA TWI NS7 pen30 0.6889 216
## 82 pRAN TWI NS3 pen30 0.6222 1918
## 83 pRAN TWI NS5 pen30 0.6444 634
## 84 pRAN TWI NS7 pen30 0.6444 216
## 85 pSNA TWI NS3 pen45 0.6889 2857
## 86 pSNA TWI NS5 pen45 0.7111 934
## 87 pSNA TWI NS7 pen45 0.6667 310
## 88 pRAN TWI NS3 pen45 0.6778 2855
## 89 pRAN TWI NS5 pen45 0.6556 934
## 90 pRAN TWI NS7 pen45 0.7111 310
## 91 pSNA LSF NS3 pen15 0.6889 985
## 92 pSNA LSF NS5 pen15 0.6000 334
## 93 pSNA LSF NS7 pen15 0.6111 121
## 94 pRAN LSF NS3 pen15 0.6556 985
## 95 pRAN LSF NS5 pen15 0.6444 334
## 96 pRAN LSF NS7 pen15 0.6778 121
## 97 pSNA LSF NS3 pen30 0.7333 1924
## 98 pSNA LSF NS5 pen30 0.6333 629
## 99 pSNA LSF NS7 pen30 0.6333 217
## 100 pRAN LSF NS3 pen30 0.6778 1922
## 101 pRAN LSF NS5 pen30 0.6333 629
## 102 pRAN LSF NS7 pen30 0.7111 217
## 103 pSNA LSF NS3 pen45 0.7222 2856
## 104 pSNA LSF NS5 pen45 0.6444 927
## 105 pSNA LSF NS7 pen45 0.6333 309
## 106 pRAN LSF NS3 pen45 0.6556 2853
## 107 pRAN LSF NS5 pen45 0.6556 927
## 108 pRAN LSF NS7 pen45 0.6889 309
## 109 pSNA VD NS3 pen15 0.3778 1040
## 110 pSNA VD NS5 pen15 0.4111 355
## 111 pSNA VD NS7 pen15 0.4556 256
## 112 pRAN VD NS3 pen15 0.6111 1039
## 113 pRAN VD NS5 pen15 0.5778 355
## 114 pRAN VD NS7 pen15 0.6111 129
## 115 pSNA VD NS3 pen30 0.5222 2043
## 116 pSNA VD NS5 pen30 0.4444 674
## 117 pSNA VD NS7 pen30 0.4000 226
## 118 pRAN VD NS3 pen30 0.6778 2042
## 119 pRAN VD NS5 pen30 0.6778 675
## 120 pRAN VD NS7 pen30 0.6444 227
## 121 pSNA VD NS3 pen45 0.5667 3035
## 122 pSNA VD NS5 pen45 0.4889 992
## 123 pSNA VD NS7 pen45 0.4000 322
## 124 pRAN VD NS3 pen45 0.6333 3037
## 125 pRAN VD NS5 pen45 0.6000 992
## 126 pRAN VD NS7 pen45 0.6444 323
## 127 pSNA RSP NS3 pen15 0.3667 1043
## 128 pSNA RSP NS5 pen15 0.3889 355
## 129 pSNA RSP NS7 pen15 0.4444 126
## 130 pRAN RSP NS3 pen15 0.6111 1043
## 131 pRAN RSP NS5 pen15 0.6000 355
## 132 pRAN RSP NS7 pen15 0.6444 127
## 133 pSNA RSP NS3 pen30 0.4556 2035
## 134 pSNA RSP NS5 pen30 0.4333 673
## 135 pSNA RSP NS7 pen30 0.4444 225
## 136 pRAN RSP NS3 pen30 0.6444 2034
## 137 pRAN RSP NS5 pen30 0.6667 674
## 138 pRAN RSP NS7 pen30 0.6333 226
## 139 pSNA RSP NS3 pen45 0.5556 3032
## 140 pSNA RSP NS5 pen45 0.4889 991
## 141 pSNA RSP NS7 pen45 0.3778 320
## 142 pRAN RSP NS3 pen45 0.6444 3034
## 143 pRAN RSP NS5 pen45 0.6556 992
## 144 pRAN RSP NS7 pen45 0.6222 321write.csv(precision_df,"结果精度_8单因子_origin.csv",row.names = F,fileEncoding="utf-8")
pdf_sna<-subset(precision_df,method=="pSNA")
pdf_ran<-subset(precision_df,method=="pRAN")
pdf_ran_mean<-aggregate(pdf_ran[,c("precisionValue","sampleNum")],by=list(method=pdf_ran$method,NS=pdf_ran$NS,percent=pdf_ran$percent),FUN=mean)
pdf_ran_mean$factor<-"Random"
precision_df_final<-rbind(pdf_sna,pdf_ran_mean)
print(precision_df_final)## method factor NS percent precisionValue sampleNum
## 1 pSNA DEM NS3 pen15 0.3222000 1040.000
## 2 pSNA DEM NS5 pen15 0.2556000 354.000
## 3 pSNA DEM NS7 pen15 0.4778000 130.000
## 7 pSNA DEM NS3 pen30 0.3778000 2040.000
## 8 pSNA DEM NS5 pen30 0.4222000 677.000
## 9 pSNA DEM NS7 pen30 0.3889000 226.000
## 13 pSNA DEM NS3 pen45 0.4667000 3036.000
## 14 pSNA DEM NS5 pen45 0.4333000 993.000
## 15 pSNA DEM NS7 pen45 0.3889000 326.000
## 19 pSNA Slope NS3 pen15 0.7333000 994.000
## 20 pSNA Slope NS5 pen15 0.6667000 336.000
## 21 pSNA Slope NS7 pen15 0.6667000 121.000
## 25 pSNA Slope NS3 pen30 0.7000000 1947.000
## 26 pSNA Slope NS5 pen30 0.6889000 636.000
## 27 pSNA Slope NS7 pen30 0.6000000 214.000
## 31 pSNA Slope NS3 pen45 0.7333000 2885.000
## 32 pSNA Slope NS5 pen45 0.6444000 929.000
## 33 pSNA Slope NS7 pen45 0.6556000 307.000
## 37 pSNA PlanC NS3 pen15 0.6333000 1011.000
## 38 pSNA PlanC NS5 pen15 0.6778000 345.000
## 39 pSNA PlanC NS7 pen15 0.5333000 129.000
## 43 pSNA PlanC NS3 pen30 0.6222000 1973.000
## 44 pSNA PlanC NS5 pen30 0.6778000 649.000
## 45 pSNA PlanC NS7 pen30 0.5222000 227.000
## 49 pSNA PlanC NS3 pen45 0.6222000 2937.000
## 50 pSNA PlanC NS5 pen45 0.6556000 953.000
## 51 pSNA PlanC NS7 pen45 0.5667000 319.000
## 55 pSNA ProfileC NS3 pen15 0.6444000 996.000
## 56 pSNA ProfileC NS5 pen15 0.5778000 338.000
## 57 pSNA ProfileC NS7 pen15 0.6222000 122.000
## 61 pSNA ProfileC NS3 pen30 0.6444000 1946.000
## 62 pSNA ProfileC NS5 pen30 0.7000000 636.000
## 63 pSNA ProfileC NS7 pen30 0.5778000 215.000
## 67 pSNA ProfileC NS3 pen45 0.6444000 2895.000
## 68 pSNA ProfileC NS5 pen45 0.6444000 934.000
## 69 pSNA ProfileC NS7 pen45 0.6333000 307.000
## 73 pSNA TWI NS3 pen15 0.7333000 983.000
## 74 pSNA TWI NS5 pen15 0.8000000 337.000
## 75 pSNA TWI NS7 pen15 0.7111000 122.000
## 79 pSNA TWI NS3 pen30 0.6889000 1919.000
## 80 pSNA TWI NS5 pen30 0.7667000 634.000
## 81 pSNA TWI NS7 pen30 0.6889000 216.000
## 85 pSNA TWI NS3 pen45 0.6889000 2857.000
## 86 pSNA TWI NS5 pen45 0.7111000 934.000
## 87 pSNA TWI NS7 pen45 0.6667000 310.000
## 91 pSNA LSF NS3 pen15 0.6889000 985.000
## 92 pSNA LSF NS5 pen15 0.6000000 334.000
## 93 pSNA LSF NS7 pen15 0.6111000 121.000
## 97 pSNA LSF NS3 pen30 0.7333000 1924.000
## 98 pSNA LSF NS5 pen30 0.6333000 629.000
## 99 pSNA LSF NS7 pen30 0.6333000 217.000
## 103 pSNA LSF NS3 pen45 0.7222000 2856.000
## 104 pSNA LSF NS5 pen45 0.6444000 927.000
## 105 pSNA LSF NS7 pen45 0.6333000 309.000
## 109 pSNA VD NS3 pen15 0.3778000 1040.000
## 110 pSNA VD NS5 pen15 0.4111000 355.000
## 111 pSNA VD NS7 pen15 0.4556000 256.000
## 115 pSNA VD NS3 pen30 0.5222000 2043.000
## 116 pSNA VD NS5 pen30 0.4444000 674.000
## 117 pSNA VD NS7 pen30 0.4000000 226.000
## 121 pSNA VD NS3 pen45 0.5667000 3035.000
## 122 pSNA VD NS5 pen45 0.4889000 992.000
## 123 pSNA VD NS7 pen45 0.4000000 322.000
## 127 pSNA RSP NS3 pen15 0.3667000 1043.000
## 128 pSNA RSP NS5 pen15 0.3889000 355.000
## 129 pSNA RSP NS7 pen15 0.4444000 126.000
## 133 pSNA RSP NS3 pen30 0.4556000 2035.000
## 134 pSNA RSP NS5 pen30 0.4333000 673.000
## 135 pSNA RSP NS7 pen30 0.4444000 225.000
## 139 pSNA RSP NS3 pen45 0.5556000 3032.000
## 140 pSNA RSP NS5 pen45 0.4889000 991.000
## 141 pSNA RSP NS7 pen45 0.3778000 320.000
## 11 pRAN Random NS3 pen15 0.6611250 1012.125
## 22 pRAN Random NS5 pen15 0.6263875 344.750
## 34 pRAN Random NS7 pen15 0.6527750 125.250
## 4 pRAN Random NS3 pen30 0.6569500 1978.125
## 5 pRAN Random NS5 pen30 0.6652750 651.875
## 6 pRAN Random NS7 pen30 0.6569250 221.125
## 71 pRAN Random NS3 pen45 0.6500000 2941.750
## 82 pRAN Random NS5 pen45 0.6514125 957.625
## 94 pRAN Random NS7 pen45 0.6763875 315.500write.csv(precision_df_final,"结果精度_8单因子_final.csv",row.names = F,fileEncoding="utf-8")